Diabetes in Old Age THIRD EDITION

Exercise and Sport in Diabetes Second Edition Edited by Dinesh Nagi 978 0470 022061

Complementary Therapies and the Management of Diabetes and Vascular Disease Edited by Patricia Dunning 978 0470 014585

Diabetes in Clinical Practice: Questions and Answers from Case Studies Edited by N. Katsilambros, E. Diakoumopoulou, I. Ioannidis, S. Liatis, K. Makrilakis, N. Tentolouris and P. Tsapogas 978 0470 035221

Obesity and Diabetes Second Edition Edited by Anthony Barnett and Sudhesh Kumar 9780 0470 848982

Prevention of Type 2 Diabetes Edited by Manfred Ganz 978 0470 857335

Diabetes – Chronic Complications Second Edition Edited by Kenneth Shaw and Michael Cummings 978 0470 865798

The Metabolic Syndrome Edited by Christopher Byrne and Sarah Wild 978 0470 025116

Psychology in Diabetes Care Second Edition Edited by Frank J. Snoek and T. Chas Skinner 978 0470 015049

The Foot in Diabetes Fourth Edition Edited by Andrew J. M. Boulton, Peter R. Cavanagh and Gerry Rayman 978 0470 015049

Gastrointestinal Function in Diabetes Mellitus Edited by Michael Horowitz and Melvin Samson 978 0471 899167

Diabetic Nephropathy Edited by Christopher Hasslacher 978 0471 589924

Nutritional Management of Diabetis Mellitus Edited by Gary Frost, Anne Dornhorst and David R. Hadden 978 0471 962045

Hypoglycaemia in Clinical Diabetes Second Edition Edited by Brian M. Frier and Miles Fisher 978 0470 018446

Diabetic Cardiology Edited by Miles Fisher and John McMurray 978 0470 862049 Diabetes in Old Age THIRD EDITION

Editor Alan J. Sinclair

Bedfordshire & Hertfordshire Postgraduate Medical School, University of Bedfordshire, UK

Wiley-Blackwell is an imprint of John Wiley & Sons, formed by the merger of Wiley’s global Scientiﬁc, Technical and Medical business with Blackwell Publishing.

Registered ofﬁce: John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK

Other Editorial Ofﬁces 9600 Garsington Road, Oxford, OX4 2DQ, UK 111 River Street, Hoboken, NJ 07030-5774, USA

For details of our global editorial ofﬁces, for customer services and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com/wiley-blackwell

The right of the author to be identiﬁed as the author of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher.

Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books.

Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought.

The contents of this work are intended to further general scientific research, understanding, and discussion only and are not intended and should not be relied upon as recommending or promoting a specific method, diagnosis, or treatment by physicians for any particular patient. The publisher and the author make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of fitness for a particular purpose. In view of ongoing research, equipment modifications, changes in governmental regulations, and the constant flow of information relating to the use of medicines, equipment, and devices, the reader is urged to review and evaluate the information provided in the package insert or instructions for each medicine, equipment, or device for, among other things, any changes in the instructions or indication of usage and for added warnings and precautions. Readers should consult with a specialist where appropriate. The fact that an organization or Website is referred to in this work as a citation and/or a potential source of further information does not mean that the author or the publisher endorses the information the organization or Website may provide or recommendations it may make. Further, readers should be aware that Internet Websites listed in this work may have changed or disappeared between when this work was written and when it is read. No warranty may be created or extended by any promotional statements for this work. Neither the publisher nor the author shall be liable for any damages arising herefrom.

ISBN: 978-0-470-06562-4

A catalogue record for this book is available from the British Library.

Typeset in 9.75/11.75 Times Roman by Laserwords Private Limited, Chennai, India Printed and bound in Great Britain by Antony Rowe Ltd, Chippenham, Wilts First Impression 2009 Contents

Foreword xi 3.3 Why detect diabetes? 23 3.4 The symptoms of diabetes 24 Preface xv 3.5 Glycosuria 25 3.6 Fasting plasma glucose and the modified oral List of Contributors xvii glucose tolerance test 25 3.7 Random and postprandial plasma glucose 27 3.8 Blood glucose meters 28 SECTION I EPIDEMIOLOGY AND 3.9 Glycosylated haemoglobin 29 PATHOPHYSIOLOGY 3.10 Fructosamine 30 3.11 Diabetes prediction calculators 31 3.12 Known diabetes 31 1 Pathophysiology of Diabetes In The Elderly 3 3.13 Types of diabetes 32 Graydon S. Meneilly 3.14 Metabolic syndrome 33 1.1 Introduction 3 3.15 Future research 33 1.2 Diet and diabetes in the elderly 4 3.16 Conclusions 33 1.3 Other factors 5 References 34 1.4 Metabolic alterations 5 1.5 Molecular biology studies 8 1.6 Glucose counter-regulation 8 SECTION III VASCULAR COMPLICATIONS 1.7 Conclusions 9 Acknowledgments 9 4 Peripheral Arterial Disease in Old People with References 9 Diabetes 41 2 Diabetes-Related Risk Factors In Leocadio Rodrıguez´ Mañas, Cristina Alonso Older People 13 Bouzon and Marta Castro Rodrıguez´ Stephen Colagiuri and Daniel Davies 4.1 Introduction 41 2.1 Introduction 13 4.2 Epidemiology of peripheral arterial disease 42 2.2 Age 13 4.3 Pathophysiology 42 2.3 Impaired glucose tolerance (IGT) and impaired 4.4 Clinical presentation 45 fasting glucose (IFG) 14 4.5 Diagnostic methods 46 2.4 Body weight 14 4.6 Treatment 47 2.5 Ethnicity 14 References 48 2.6 Gestational diabetes mellitus (GDM) 14 2.7 Family history 14 5 Coronary Heart Disease 51 2.8 Hypertension 15 2.9 Cardiovascular and cerebrovascular disease 15 Ahmed H. Abdelhafiz 2.10 Physical inactivity 15 5.1 Introduction 51 2.11 Antipsychotic medication and Mental Illness 15 5.2 Effect of ageing and diabetes on the 2.12 Sleep disorders 16 cardiovascular system 51 2.13 Smoking 16 5.3 Epidemiology 53 2.14 Conclusions 16 5.4 Cardiovascular risks 53 References 16 5.5 Myocardial infarction 61 5.6 Heart failure 63 5.7 Anti-diabetic medications and CHD 64 SECTION II SCREENING AND 5.8 Conclusions 66 DIAGNOSIS References 66 3 Diabetes in the Elderly: Diagnosis, Testing 6 Stroke and Diabetes 75 and Screening 21 Janice E. O’Connell and Christopher S. Gray Simon Croxson 6.1 Introduction 75 3.1 Introduction 21 6.2 Diabetes, hyperglycaemia and stroke risk in older 3.2 Definition of diabetes in the elderly 21 people 76 vi CONTENTS

6.3 Diabetes, post-stroke hyperglycaemia and 10.9 Pressure palsies 141 prognosis after acute stroke 79 10.10 Pathogenesis of diabetic peripheral neuropathy 142 6.4 Diabetes, hyperglycaemia and acute stroke 10.11 Autonomic neuropathy 142 treatment 80 10.12 Management of painful diabetic neuropathy 144 6.5 Stroke prevention in type 2 diabetes 84 References 146 6.6 Conclusions 87 References 88 11 Erectile Dysfunction 149 Tamás Várkonyi and Peter Kempler 7 Diabetes-Related Renal Disease in Older 11.1 Introduction 149 People 93 11.2 Erectile dysfunction: An observable marker of Latana A. Munang and John M. Starr diabetes mellitus? 150 7.1 Introduction 93 11.3 Risk factors of erectile dysfunction 150 7.2 Changes in the diabetic kidney 93 11.4 Pathophysiology of ED in diabetes mellitus 151 7.3 Screening for diabetic renal disease 94 11.5 Erectile dysfunction: A first sign of 7.4 Chronic kidney disease 95 cardiovascular disease? 151 7.5 Management of diabetic renal disease 95 11.6 The artery size hypothesis: A macrovascular 7.6 Managing cardiovascular risk 97 link between ED and CAD 152 7.7 Management of CKD and its complications 97 11.7 Other causes of erectile dysfunction 152 7.8 Conservative management of renal disease 98 11.8 Differentiation between organic and 7.9 Conclusions 99 psychogenic erectile dysfunction 153 References 100 11.9 Treatment of erectile dysfunction in diabetic patients 153 11.10 Conclusions 158 8 Management of Eye Disease and Visual Loss 103 References 158 Nina Tumosa 8.1 Introduction 103 8.2 Risk factors 103 SECTION IV TREATMENT AND CARE ISSUES 8.3 Management of diabetic retinopathy 104 8.4 Treatments of diabetic retinopathy 106 12 Metabolic Risk Factors, Obesity and 8.5 Conclusions 109 Cardiometabolic Syndrome 165 References 110 Raffaele Marfella and Giuseppe Paolisso 9 The Diabetic Foot 113 12.1 The definition of metabolic syndrome 165 12.2 Pathogenesis of the metabolic syndrome 167 Andrew J. M. Boulton and Matthew J. Young 12.3 Metabolic syndrome in older persons 168 9.1 Introduction 113 12.4 Insulin resistance and dyslipidaemia 168 9.2 Peripheral sensorimotor neuropathy 114 12.5 Insulin resistance and obesity 169 9.3 Autonomic neuropathy 116 12.6 Insulin resistance and arterial hypertension 170 9.4 Peripheral vascular disease 117 12.7 Potential components of the metabolic 9.5 Limited joint mobility 118 syndrome in the elderly 171 9.6 Foot pressure abnormalities 118 12.8 Therapeutic perspectives 173 9.7 Other risk factors 119 12.9 Conclusions 176 9.8 The classification of ulceration 120 References 176 9.9 The At-risk foot 120 9.10 Superficial ulcers: Wagner/UT Grade 1 121 13 Perspectives on Diabetes Care in Old Age: A 9.11 Deep ulcers: Wagner/UT Grades 2 and 3 124 Focus on Frailty 183 9.12 Localized gangrene: Wagner Grade 4 126 9.13 Extensive gangrene: Wagner Grade 5 129 Andrej Zeyfang and Jeremy D. Walston 9.14 The diabetic Charcot foot 129 13.1 Introduction 183 9.15 Conclusions 131 13.2 The frailty syndrome and its biology 184 References 131 13.3 Frailty and diabetes in late life 184 13.4 The biology of late-life diabetes 185 10 Diabetic Neuropathy 137 13.5 Geriatric syndromes and diabetes 187 13.6 Approaches for improving care for frail, older Solomon Tesfaye diabetics 188 10.1 Introduction 137 13.7 Drug therapy in the frail elderly 189 10.2 Epidemiology 137 13.8 Summary 191 10.3 Classification of diabetic polyneuropathy 137 References 191 10.4 Diabetic peripheral neuropathy (DPN) 137 10.5 Asymmetrical neuropathies 140 14 Metabolic Decompensation in the Elderly 195 10.6 Diabetic amyotrophy 140 10.7 Cranial mononeuropathies 141 Giuseppe Paolisso and Michelangela Barbieri 10.8 Thoracoabdominal neuropathy 141 14.1 Introduction 195 CONTENTS vii

14.2 Hypoglycaemia 195 18.1 Introduction 265 14.3 Diabetic ketoacidosis and hyperosmolar 18.2 What glycaemic threshold should trigger insulin hyperglycaemic state in the elderly 198 therapy in the elderly? 266 14.4 Pathogenesis of DKA and HHS 199 18.3 Other indications for insulin therapy in the 14.5 Diagnosis of DKA and HHS 200 elderly 266 14.6 Treatment of DKA and HHS 201 18.4 Some of special considerations of elderly people 14.7 Treatment-related complications 204 and insulin therapy 267 14.8 Conclusions 205 18.5 Glucose-lowering medications and the References 205 introduction of insulin therapy 268 18.6 Which insulin regimen should be used in the elderly? 269 15 Nutritional Perspectives: Diabetes in Older 18.7 Titrating the insulin dosage and monitoring People 209 progress 270 18.8 What should patients be told on the day Begoña Molina and Alan J. Sinclair they start insulin? 270 15.1 Introduction 209 18.9 When should more complex insulin regimens be 15.2 Basis of nutritional support 210 introduced? 271 15.3 Nutritional assessment 214 18.10 Favourable outcomes 271 15.4 Energy intake: carbohydrates and fats 218 References 271 15.5 Nutritional oral supplements 223 15.6 Artificial nutrition 223 19 Treatment of Hypertension 273 15.7 Enteral tube feeding [43, 69] 225 Peter Fasching 15.8 Parenteral nutrition [43, 69] 227 15.9 A specific nutrition support formula for 19.1 Introduction 273 elderly patients? 228 19.2 Risk assessment from observational studies 273 15.10 Ethical issues 228 19.3 Target values for hypertension treatment 15.11 Conclusions 229 in patients with diabetes 274 References 229 19.4 Available evidence from randomized controlled trials 275 19.5 Antihypertensive therapy as SECTION of a 16 Early Management of Type 2 Diabetes 233 multifactorial intervention 278 19.6 Pharmacological therapy of hypertension in Alan Sinclair diabetes 278 16.1 Introduction 233 References 278 16.2 Developing the case for high-quality diabetes care 234 20 Treatment of Dyslipidaemia 281 16.3 Aims in the early management 235 16.4 Initial treatment 237 Peter Fasching 16.5 Establishing an individual diabetes care plan 241 20.1 Introduction 281 16.6 Conclusions 242 20.2 Risk assessment and treatment targets 281 References 242 20.3 Available evidence from randomized controlled trials 282 20.4 Eligibility for lipid-lowering therapy 284 17 Drug Therapy: Current and Emerging Agents References 284 for Hyperglycaemia 245 Joe M. Chehade and Arshag D. Mooradian 21 Hypoglycaemia 287 17.1 Introduction 245 Vincent McAulay and Brian M. Frier 17.2 Sulphonylureas 246 21.1 Introduction 287 17.3 Biguanides 248 21.2 Physiological responses to hypoglycaemia 288 17.4 Alpha-glucosidase inhibitors 249 21.3 Acquired hypoglycaemia syndromes in type 1 17.5 Thiazolidinediones 250 diabetes 290 17.6 Meglitinides 251 21.4 Effects of age on physiological responses to 17.7 Incretin mimetics and enhancers: Glucagon-like hypoglycaemia 292 peptide-1 and dipeptidyl peptidase IV inhibitors 252 21.5 Effects of type 2 diabetes on responses to 17.8 Amylin analogues 253 hypoglycaemia 294 17.9 Insulin and insulin analogues 255 21.6 Epidemiology of hypoglycaemia in elderly 17.10 The choice of an anti-diabetic agent in elderly people with diabetes 296 people 257 21.7 Adverse effects of hypoglycaemia in the elderly 299 References 259 21.8 Risk factors for hypoglycaemia in the elderly 300 21.9 Treatment 302 21.10 Knowledge of symptoms 304 18 Insulin Therapy 265 21.11 Prevention of hypoglycaemia 304 Dennis Yue and Marg McGill 21.12 Summary and conclusions 305 References 305 viii CONTENTS

22 Diabetes in Care Homes 311 25 Managing Surgery in the Elderly Diabetic Alan Sinclair and Terry Aspray Patient 349 22.1 Introduction 311 Geoffrey Gill and Susan Benbow 22.2 The United Kingdom as a model of care home 25.1 Introduction 349 reform 311 25.2 Metabolic and other problems induced by 22.3 Epidemiology 312 surgery 349 22.4 Complications and comorbidity 313 25.3 Implications for management of surgery in 22.5 Common management problems 314 diabetic patients 350 22.6 Organization of diabetes care in residential 25.4 Potential risks of surgery in diabetic patients 351 settings 315 25.5 Special problems in the elderly 352 22.7 Sustaining effective diabetes care 321 25.6 New diabetes therapies and surgical 22.8 Conclusions 322 management 352 References 322 25.7 The practical management of surgery in diabetic patients 353 25.8 Conclusions 358 23 Primary and Community Care of Diabetes in References 358 Older People 325 Roger Gadsby SECTION V MANAGEMENT OF 23.1 Introduction 325 23.2 The shift of diabetes care from the hospital to the ASSOCIATED community 326 COMPLICATIONS 23.3 The primary care diabetes team 327 23.4 Routine care of people with diabetes by the 26 The Implications of the Evolving Diabetes primary care diabetes team 327 Epidemic for Disability in Older Adults 363 23.5 Which strategies to prevent CVD are appropriate Edward W. Gregg and Linda Geiss for older people with diabetes? 328 23.6 Preventive health care in older people: Further 26.1 Diabetes and older adults: The magnitude and debate 331 character of the problem 363 23.7 Delivering appropriate diabetes care 331 26.2 Diabetes and functional disability 365 23.8 Problems of polypharmacy and medication 26.3 Modiﬁable factors explaining diabetes and concordance in older people with diabetes 333 disability 366 23.9 Conclusions 334 26.4 Interventions to reduce disability risk 367 References 334 26.5 Trends in disability among the diabetic population 368 26.6 Summary 368 References 369 24 Diabetes Care in Special Circumstances 337 27 Diabetes and Cognitive Dysfunction 373 I Acute Hospital Admission: Jay Chillala Alan J. Sinclair and Koula G. Asimakopoulou 27.1 Introduction 373 24.1 Introduction 337 27.2 Evidence of association between diabetes and 24.2 Hyperosmolar hyperglycaemic state and cognitive dysfunction 374 diabetic ketoacidosis insulin regimes 337 27.3 Cognitive dysfunction and glycaemic control 376 24.3 Factors which affect management 338 27.4 The importance of detecting cognitive 24.4 Discharge planning 338 dysfunction 377 II Serious Infections: Jay Chillala 27.5 The importance of excluding depression 380 27.6 Further investigations 380 24.5 Introduction 340 27.7 Conclusions 380 24.6 How is the treatment of elderly patients different? 340 References 381 24.7 Management 341 24.8 Precautions for the medical management of infections in the elderly 341 28 Depression and Diabetes in Older Adults 385 24.9 When should the microbiologist be called? 342 Arie Nouwen and Jan R. Oyebode

III Minor Ethnic Populations: Gurch Randhawa 28.1 Deﬁning depression 385 28.2 Depression in older people 387 24.10 Introduction 344 28.3 Epidemiology of depression in diabetes 387 24.11 Epidemiology of diabetes among minority ethnic 28.4 Depression in diabetes: Associated problems 388 groups in the UK 344 28.5 Possible mechanisms for the aetiology of 24.12 Improving access to services 345 depression 389 24.13 Looking to the future 345 28.6 The assessment of depression in older adults 24.14 Conclusions 346 with diabetes 390 References 346 28.7 The treatment of clinical depression 393 CONTENTS ix

28.8 Case studies 395 31.9 Conclusions 441 28.9 Summary 397 31.10 Acknowledgments 441 References 397 References 441

29 Diabetes and Falls 403 32 Supporting the Family and Informal Carers 443 Christine T. Cigolle and Caroline S. Blaum Antony Bayer and Alan Sinclair 29.1 Introduction 403 32.1 Introduction 443 29.2 Deﬁnition 403 32.2 Who are the carers? 444 29.3 Epidemiology 404 32.3 What do carers do? 445 29.4 Pathophysiology of, and risk factors for, falling 405 32.4 What effect can caring have on the carer? 447 29.5 Risk factors for fall injury 408 32.5 What do carers want? 448 29.6 Evaluation of older adults with diabetes who fall 408 32.6 What are the beneﬁts of carer intervention? 449 29.7 Management of older adults with diabetes who References 449 fall 410 29.8 Patient safety and quality of care 413 33 Diabetes Mellitus Care Models for Older References 413 People 453

I The European Perspective: SECTION VI OPTIMIZING DIABETES CARE Isabelle Bourdel-Marchasson IN OLDER PEOPLE 33.1 Introduction 453 33.2 Current clinical guidelines in Europe 454 30 The Role of Specialist Nurses and Other Members 33.3 Care organization 456 of the Multidisciplinary Team (MDT) 419 33.4 Call for audit of quality of care 457 Carolin D. Taylor and Timothy J. Hendra 33.5 Perspectives 458 30.1 Introduction 419 II The United States Perspective: 30.2 The relationship of the diabetes MDT with other John E. Morley services for older people 420 30.3 The changing role of the MDT in service delivery 420 33.6 Introduction 459 30.4 The ‘‘Patient’’ 420 33.7 Speciﬁc issues in older diabetics 459 30.5 The roles of the individual team members 422 33.8 Management of diabetes 460 30.6 The role of the MDT in speciﬁc situations 424 33.9 Conclusions 461 30.7 Summary 425 Acknowledgments 462 References 425 References 462

31 Diabetes Education in the Elderly 427 34 Further Initiatives to Enhance Diabetes Care Charles Fox and Anne Kilvert in Older People 465 31.1 Introduction 427 Alan Sinclair 31.2 Principles of adult diabetes education 428 34.1 Introduction 465 31.3 How should education be delivered? 428 34.2 Improving inpatient diabetes care 466 31.4 Empowerment in diabetes care 429 34.3 The use of a diabetes minimum dataset (MDS) to 31.5 The development of structured education: What to enhance care 468 teach people about diabetes 434 34.4 Conclusions 471 31.6 Adaptations of educational principles for the References 471 elderly 437 31.7 Assessment of the educational needs of the individual 439 Index 473 31.8 Speciﬁc cases 439 Foreword

The cognoscenti, the small cadre of experts on diabetes swimmers, dancers and authors, striving for excellence in the elderly, will skip this foreword and dive right channels energies and rejuvenates the self. The physi- into the individual chapters. There they will find many cian who adopts the mindset of a champion helps his treasures related to clinical science and clinical care, or her patients, helps other health care professionals as well as historical vignettes and current controversies with their patients and nourishes his or her own soul. related to diabetes in the older patient. At this time in medicine, when physician burnout is You, by reading this foreword in a book on dia- epidemic, nourishment for the soul can be life-saving. betes in the elderly patient, are marking yourselves as In the USA, where the pension systems are in non-expert but you are clearly ahead of your medical disarray and large debts have been piled up to pay colleagues. You are recognizing that the excellent text- for schooling, physicians will be working many years books on diabetes and excellent textbooks on geriatric past the hallowed 65. The best preparation for the medicine, though they cover medical care of the older long journey is passion in one’s professional pursuit. patient, typically fall short in dealing with the older As an internist, or endocrinologist, or geriatrician, join patient with diabetes. me in exploring the attractions of becoming skilled in These textbooks mirror the state of affairs in medi- the care of diabetes in the elderly. cal care today. When I was a young physician, I was When I entered the profession fifty years ago, anti- impressed that excellent internists provided excellent biotics were routing many infectious diseases. The care for their patients, including very good diabetes ancient aphorism “If you know syphilis, you know management. My impression now is that very good in- all of medicine” was being re-modelled; syphilis was ternists continue to provide very good care, except for replaced by diabetes. diabetes where the care often is only mediocre. Many I propose a new model: “If you know diabetes in endocrinologists, formerly excellent in diabetes, are the elderly, you know all of medicine”. also falling further and further back from the cutting edge of diabetes care. This is especially sad because The challenge for the profession we now know more than ever the importance of good Increasingly, medicine in general is benefiting from the management and have better tools with which to ap- introduction of protocols and algorithms. While im- proach the desired goals. The gap between “excellent” proving care, these also shrink the intellectual distance and “actual” widens as the patient’s age increases. between the physician, the physician’s assistant and In this essay, I plan to inspire you, to help guide the nurse. I am guessing that a 37-year-old professor you into a highly satisfying professional path, a path of computer science with type 1 diabetes can prob- that will please you, as well as enhance your value ably manage well with a little help from a diabetes to your patients and to your medical community. The educator and an occasional visit to a physician. Re- rest of this book is filled with instructional material call the World War II pharmacist’s mate who in the that you will find very useful. My goal is to provide pre-antibiotic area successfully removed an inflamed an overarching view from the top of the mountain. appendix from a crew member of his submarine sub- merged beneath the waters of the Pacific. Nourishing the soul Advancing age brings growing complexity. Elderly Champions seek new challenges, set new goals. patients with diabetes need continuous input from For mountain climbers and cellists, surgeons and skilled physicians. For these physicians, protocols and xii FOREWORD algorithms are the starting point but the real plan needs (i) depression without sadness is easy to miss. multiple modifications, surveillance, balancing of com- (ii) Screening instruments are helpful. peting priorities, and skilled navigation of poorly charted waters. It demands professional skills at their (iii) Personalized rationalizations (“If I was 82 and best. living alone, I would also feel that way ... ”) can obscure the correct diagnosis and management. Interpreting data (iv) Drugs as well as endocrine diseases and other Multi-centre trials, the foundation of therapeutics to- disorders are common aetiologies of depression day, are typically performed on younger patients. With that is reversible. the basic and clinical science in the background, the (v) When medication and psychotherapy fail, ECT data from widely heralded multi-centre trials (with pa- (electroconvulsive therapy), is an excellent thera- tients who are typically younger and less complicated) peutic choice to consider. provide a basis but not a recipe for care of the elderly patient. Advanced age and other exclusionary criteria, (vi) With ageing, suicide rates rise sharply, especially including medications, make extrapolations to the el- among white males. derly more tenuous. The loud “microphones” supported by pharmaceutical company coffers often fill the air with information that is misleading for older patients. Demographics and disease Laboratory standards are based on younger popula- The population is being enriched progressively with tions. Data in the elderly are much sparser. Even when patients who are over 65. They are living longer. the mean and median for a lab test remain unchanged, The so-called old-old are a rapidly growing group. the splay typically increases so that higher and lower Objective data to guide the physician require ever values that are “normal” for an older patient are easily longer lines of extrapolation, demanding more of the labelled as pathological. physician’s judgment. The incidence and prevalence New medications are largely tested on younger, less of diabetes increases with age. Ageing brings out complicated patients. Data among older patients are diabetes, and diabetes accelerates biological ageing sparse. Many side effects of drugs emerge gradually and onset of other pathology. Both processes corrode in the years after their introduction. The catalogue of cognition. side effects among older complex patients emerge more Ageing in our Society: The universal reverence, or at slowly. The sparseness of data dictates that new drugs least respect, for the elderly that held sway worldwide should be avoided in older patients, except on the very since the beginning of human memory, has been re- rare occasion when the new drug is a very substantial placed in the industrialized world of today with a wide advance and other drugs cannot meet the need. range of negative attitudes, mostly undeserved. In their Adverse drug interactions between two drugs are care for the elderly, physicians and their teammates in identified slowly. Many remain undetected. Typical care will be energized by recalling the widely appreci- elderly patients take many medications, exponentially ated positive features of a majority of the elderly: increasing the likelihood of adverse drug interactions and, equally, making their detection most difficult. (i) Every elderly patient can be improved in some way by an encounter with a professional. Depression (ii) Typically, the elderly are appreciative of the care and express their appreciation. Advancing age as well as medications and multiple medical conditions are associated with depression. The (iii) Their expectations for improvement are realisti- link between diabetes and depression has received a lot cally tempered. of attention recently. Growing evidence that depression (iv) They are individually “more unique”. impacts negatively on physical health mandates that depression, so common in the elderly, be detected and “More unique” is a phrase that will galvanize to treated energetically. action legions of amateur grammarians all over the In dealing with depression, especially in the older English-speaking world. They will reflexly remind me patient, recall that that unique indicates one-of-a-kind and therefore no FOREWORD xiii comparator is permitted. Biology and I will prove them Valediction wrong. Let’s start with a fertilized egg that is just dividing to generate a pair of monozygotic twins. They With a little luck, it is likely that you, in your lifetime, are not identical and progressively diverge, distanc- will never lack for food for your body. Much more at ing one biological self from the other. All humans risk, and therefore more to be guarded, is the supply do the same. The extremely similar looking zygotes, of nourishment for your professional soul. and highly similar looking newborns progressively di- Jesse Roth MD verge, biologically, sociologically and medically, to the Feinstein Institute for Medical Research, delight and amazement of the skilled physician and North Shore-LIJ Health System and other health care providers. Like snowflakes, Rem- Albert Einstein College of Medicine, brandt paintings, precious gemstones, and leaves from Yeshiva University a single tree, blessedly, there are no sames. New York Preface

In this Third Edition, I have assembled an international diabetes care team, and in this book we offer direction set of distinguished authors from a range of medi- in good clinical practice and a template for assessment cal disciplines to provide insight and expertise in the and further care. We hope that this book will continue complex environment of the management of older peo- to stimulate interest in geriatric diabetes and that, with ple with diabetes. As before, this book should provide the publication of further large clinical trials in this up-to-date knowledge and guidance in effective clini- area, the evidence base will be enriched. cal decision-making both for the generalist as well as for the specialist. Unlike other books about geriatric Acknowledgements medical practice, our authors have designed their individual chapters around the unique characteristics of This book is dedicated to all those who strive to older people with diabetes rather than merely adding enhance the quality of diabetes care for older people. one or two speciﬁc paragraphs at the end of a more Very special thanks to my family and to many general chapter. helpful colleagues at John Wiley. It is increasingly recognised that very elderly peo- Alan J Sinclair ple with diabetes pose considerable challenges to the March 2009 List of Contributors

Ahmed H. Abdelhafiz Susan Benbow Department of Elderly Medicine Department of Diabetes and Endocrinology Rotherham General Hospital Aintree University Hospitals Moorgate Road Liverpool L9 1AE Rotherham United Kingdom S60 2UD United Kingdom Caroline S. Blaum Department of Internal Medicine Koula G. Asimakopoulou University of Michigan King’s College London Ann Arbor Dental Institute Michigan 48109 Oral Health Services Research and Dental Public USA Health Denmark Hill and London SE5 9RW United Kingdom Ann Arbor VA Healthcare System GRECC Terry Aspray Michigan 48109 Sunderland Royal Hospital USA Kayll Road Sunderland SR4 7TP Andrew J. M. Boulton Manchester Royal Infirmary and Oxford Road Newcastle University Manchester M13 9WL Campus for Ageing and Vitality United Kingdom Newcastle upon Tyne NE4 5PL Isabelle Bourdel-Marchasson United Kingdom Geriatric Department Michelangela Barbieri Hôpital Xavier Arnozan Department of Geriatric Medicine and CHU of Bordeaux Metabolic Diseases 33604 Pessac cedex Second University of Naples (SUN) UMR 5536 CNRS/Université Victor Segalen Piazza Miraglia Bordeaux 2 2 80138 Naples 146 rue Léo Saignat Italy 33000 Bordeaux France Antony Bayer Department of Geriatric Medicine Cristina Alonso Bouzon Academic Centre Servicio de Geriatr´ıa Cardiff University Hospital Universitario de Getafe Llandough Hospital Ctra. de Toledo, Km. 12,5. Penarth 28905-Getafe Vale of Glamorgan CF64 2XX Madrid United Kingdom Spain xviii LIST OF CONTRIBUTORS

Joe M. Chehade Charles Fox Division of Endocrinology, Diabetes and Metabolism Diabetes Centre University of Florida College of Medicine Northampton General Hospital Jacksonville 51 The Avenue FL 32209 Northampton NN1 5BT USA United Kingdom

Jay Chillala Brian M. Frier Trafford General Hospital Department of Diabetes Moorside Road Royal Inﬁrmary Davyhulme 51 Little France Crescent Manchester M41 5SL Edinburgh EH16 4SA United Kingdom United Kingdom

Christine T. Cigolle Roger Gadsby Department of Family Medicine Warwick Medical School University of Michigan University of Warwick Ann Arbor Gibbet Hill Road Michigan 48109 Coventry CV4 7AL USA United Kingdom and Linda Geiss Division of Diabetes Translation Ann Arbor VA Healthcare System Centers for Disease Control and Prevention GRECC 4770 Buford Highway Michigan 48109 N.E. Mailstop K-10 USA Atlanta GA USA Stephen Colagiuri Institute of Obesity Nutrition and Exercise Geoffrey Gill K25 - Medical Foundation Building Department of Diabetes and Endocrinology The University of Sydney Aintree University Hospitals NSW 2006 Liverpool L9 1AE Australia United Kingdom

Simon Croxson Christopher S. Gray United Bristol Healthcare NHS Trust Department of Geriatric Medicine Bristol General Hospital University of Newcastle Bristol BS1 6SY Newcastle on Tyne United Kingdom United Kingdom

Daniel Davies Edward W. Gregg Institute of Obesity Nutrition and Exercise Division of Diabetes Translation K25 - Medical Foundation Building Centers for Disease Control and Prevention The University of Sydney 4770 Buford Highway NSW 2006 N.E. Mailstop K-10 Australia Atlanta GA USA Peter Fasching 3rd International Department PH, Timothy J. Hendra Baumgarten, Department of Geriatric Medicine A-1140 Vienna Robert Hadﬁeld Wing Austria Northern General Hospital LIST OF CONTRIBUTORS xix

Herries Road Begona˜ Molina Shefﬁeld S5 7AU Servicio de Endocrinolog´ıayNutrici´on United Kingdom Hospital Universitario de Getafe Carretera de Toledo Km 12,500 Peter Kempler 28905-Getafe I. Department of Medicine Madrid Semmelweis University Spain Budapest Hungary Arshag D. Mooradian Division of Endocrinology, Diabetes and Anne Kilvert Metabolism Diabetes Centre University of Florida College of Northampton General Hospital Medicine 51 The Avenue Jacksonville Northampton NN1 5BT FL 32209 United Kingdom USA

Leocadio Rodr´ıguez Manas˜ John E. Morley Servicio de Geriatr´ıa GRECC Hospital Universitario de Getafe VA Medical Center and Division of Geriatric Ctra. de Toledo, Km. 12,5. Medicine 28905-Getafe Saint Louis University School of Medicine Madrid 1402 S. Grand Blvd Spain M238 St Louis MO 63104 Raffaele Marfella USA Department of Geriatric and Metabolic Disease Second University Naples Latana A. Munang Piazza Miraglia Liberton Hospital 2 80138 Naples 113 Lasswade Road Italy Edinburgh EH16 6UB United Kingdom Vincent McAulay Department of Diabetes Arie Nouwen Crosshouse Hospital School of Psychology Kilmarnock KA2 0BE University of Birmingham United Kingdom Edgbaston, Birmingham B15 2TT Marg McGill United Kingdom Diabetes Centre Royal Prince Alfred Hospital Janice E. O’Connell Missenden Road Sunderland Royal Hospital Camperdown Kayll Road New South Wales Sunderland SR4 7TP Australia United Kingdom

Graydon S. Meneilly Jan R. Oyebode Division of Geriatric Medicine School of Psychology Department of Medicine University of Birmingham The University of British Columbia Edgbaston, Birmingham Vancouver BC B15 2TT Canada United Kingdom xx LIST OF CONTRIBUTORS

Giuseppe Paolisso Nina Tumosa Department of Geriatric Medicine and Metabolic Geriatrics Research, Education, and Clinical Center Diseases St. Louis VA Second University of Naples (SUN) Piazza Miraglia and 2 80138 Naples Division of Geriatrics Italy Saint Louis University Gurch Randhawa St Louis Institute for Health Research MO 63125 University of Bedfordshire USA Putteridge Bury Tamas´ Varkonyi´ Luton LU2 8LE Department of Medicine United Kingdom University of Szeged Marta Castro Rodr´ıguez Szeged Servicio de Geriatr´ıa Hungary Hospital Universitario de Getafe Jeremy D. Walston Ctra. de Toledo, Km. 12,5. Johns Hopkins University School of Medicine 28905-Getafe 5501 Hopkins Bayview Circle Madrid 21224 Baltimore Spain Maryland Alan J. Sinclair USA Bedfordshire & Hertfordshire Postgraduate Matthew J. Young Medical School Edinburgh Royal Infirmary Putteridge Bury campus 51 Little France Crescent Hitchin Road Dalkeith Road Luton LU2 8LE Edinburgh EH16 4SA United Kingdom United Kingdom John M. Starr Dennis Yue Royal Victoria Hospital Diabetes Centre Craigleith Road Royal Prince Alfred Hospital Edinburgh EH4 2DN Missenden Road United Kingdom Camperdown Carolin D. Taylor New South Wales Department of Geriatric Medicine Australia Robert Hadfield Wing Andrej Zeyfang Northern General Hospital Bethesda Hospital Stuttgart Herries Road Department of internal medicine and geriatrics Sheffield S5 7AU Hohenheimer Strasse 21 United Kingdom 70184 Stuttgart Solomon Tesfaye Germany Royal Hallamshire Hospital Sheffield Teaching Hospitals NHS Foundation Trust Glossop Road Sheffield, S10 2JF United Kingdom SECTION I Epidemiology and Pathophysiology

Graydon S. Meneilly Division of Geriatric Medicine, Department of Medicine, The University of British Columbia, Vancouver BC, Canada

Age-related Key messages decreased insulin • Lifestyle factors play a major role in diabetes in secretion the elderly. Coexisting illness • Diabetes in the elderly is metabolically distinct. • Factors Elderly patients with diabetes have an increase predisposing Age-related incidence of severe or fatal hypoglycemia. the insulin Genetics olderly resistance to diabetes 1.1 Introduction Adiposity

Drugs In the past, numerous studies have been conducted Decreased to investigate the pathogenesis of type 2 diabetes [1]. physical Although, unfortunately, elderly patients were system- activity atically excluded from these studies, we have more Figure 1.1 Factors that contribute to the high preva- recently started to study, in systematic fashion, the lence of diabetes in the elderly. Reproduced with per- pathophysiological alterations that occur in elderly pa- mission from Halter, J.B., Carbohydrate metabolism, in: tients with diabetes. These studies, the details of which E.J. Masoro (ed.), Handbook of Physiology, Volume on will be reviewed in the following sections, suggest that Aging. New York, Oxford University Press Inc., 1995, there are many ways in which diabetes in the elderly is p. 119. unique. Some of the factors that contribute to the high prevalence diabetes in the elderly are shown schemat- elderly, although the speciﬁc genes responsible have ically in Figure 1.1. yet to be deﬁned [2]. If you have a family history of type 2 diabetes, you are much more likely to develop the disease as you age [3]. Diabetes is much more 1.1.1 Genetic factors common in the elderly in certain ethnic groups [4], There are several lines of evidence which suggest that while the likelihood that an elderly identical twin will there is a strong genetic component to diabetes in the develop diabetes if their sibling is affected is over 80%.

Diabetes in Old Age Third Edition Edited by Alan J. Sinclair © 2009 John Wiley & Sons Ltd. ISBN: 978-0-470-06562-4 4 CH 01 PATHOPHYSIOLOGY OF DIABETES IN THE ELDERLY Even in elderly identical twins discordant for type 2 old age. Many older people have coexisting illnesses diabetes, the unaffected siblings clearly have evidence and take multiple drugs (e.g., thiazide diuretics), which of abnormal glucose metabolism [5]. can allow a latent abnormality in glucose metabolism to develop into full-blown diabetes [11, 12]. Obesity, 1.1.2 Age-related changes in carbohydrate especially with a central distribution of body fat, and a metabolism reduction in physical activity occur progressively with aging, and both of these factors are associated with The progressive alterations in glucose metabolism that occur with age explain why genetically susceptible abnormal carbohydrate metabolism [12–17]. older individuals may not develop diabetes until late The above information suggests that lifestyle mod- in life. Pathogenic mechanisms which contribute to ifications may be of value in the prevention of type 2 the glucose intolerance of aging include alterations diabetes in the elderly, even in patients with a strong in glucose-induced insulin release and resistance to family history of the disease. Indeed, the Diabetes insulin-mediated glucose disposal [6]. The results of Prevention Program found that a combined lifestyle early investigations suggested that glucose-induced in- intervention consisting of weight loss and increased sulin release was normal in the elderly. However, physical activity was effective in reducing the inci- more recent studies enrolling large numbers of care- dence of diabetes in elderly patients with impaired fully characterized healthy young and old subjects have glucose tolerance [18]. demonstrated definable alterations in glucose-induced insulin release in the aged [6, 7]. Of note, the magnitude of the decrement in insulin secretion is more 1.2 Diet and diabetes in the elderly apparent in response to oral than to intravenous glucose [6]. This may be due, in part, to a decreased The results of large epidemiologic studies have beta cell response to the incretin hormones (see be- shown that diabetes is more likely to develop in older low). As with many hormones, insulin is secreted patients who have a diet that is high in saturated fats in a pulsatile fashion. Normal aging is also associ- and simple sugars and low in complex carbohydrates ated with subtle alterations in pulsatile insulin release, [13, 19–21]. Moderate alcohol consumption may which further contribute to age-related changes in glu- protect against diabetes in elderly women [22]. It has cose metabolism [8]. Elevated levels of proinsulin, been suggested that deficiencies of trace elements which suggest disordered insulin processing, predict or vitamins may contribute to the development or the subsequent development of type 2 diabetes in el- progression of diabetes in younger subjects, and it is derly subjects [9]. Thus, it is clear that alterations in increasingly recognized that the same may be true in glucose-induced insulin release are an important com- the elderly [12, 21]. Elderly patients with diabetes have ponent of the changes in carbohydrate metabolism with exaggerated free radical production, and administration aging. However, the most important pathogenic mech- of the antioxidant vitamins C and E to these patients anism underlying the glucose intolerance of aging is improves both insulin action and metabolic control resistance to insulin-mediated glucose disposal [6, 10]. [23, 24]. Many elderly patients with diabetes are Debate persists as to whether the insulin resistance of deficient in magnesium and zinc, and supplements of the elderly is intrinsic to the aging process itself, or zinc and magnesium can improve glucose metabolism is the result of lifestyle factors commonly associated in these patients [25, 26]. Increased dietary iron may with aging. The consensus of opinion is that the aging be associated with an increased risk of diabetes in aged process itself is the most important cause of insulin individuals [20, 27]. Although chromium deficiency resistance, although lifestyle changes (see below), are has been shown to cause abnormalities in glucose clearly an important contributing factor. metabolism in animals and younger patients, there is no evidence to date that chromium supplements will 1.1.3 Lifestyle and environmental factors improve glucose tolerance in the elderly. In summary, Despite the strong genetic component, it is abundantly there is increasing evidence to suggest that dietary clear that various environmental and lifestyle factors abnormalities may contribute to the pathogenesis of can increase or decrease the likelihood that a geneti- diabetes in the elderly, and that dietary modifications cally susceptible individual will develop the disease in may be of therapeutic benefit in these patients. 1.4 METABOLIC ALTERATIONS 5

4 1.3 Other factors Type 2 diabetes Control The presence of inﬂammation, as evidenced by levels of proinﬂammatory cytokines such as tumour necro- ) sis factor-α (TNF-α) and C-reactive protein (CRP), 1 − is associated with an increased risk of diabetes in min 1 the elderly [28–31]. Higher levels of adiponectin (an − 3 adipocytokine that increases insulin sensitivity) are associated with a reduced incidence of diabetes in the (mg kg aged [29, 32]. Sex steroid hormone levels also appear to be related to the development of diabetes in the

elderly [33, 34]. In particular, higher testosterone lev- Fasting hepatic glucose production els in women and lower levels in men appear to be 2 associated with an increased incidence of diabetes. 46810 12 14 Fasting glucose (mmol l−1)

1.4 Metabolic alterations Figure 1.2 Fasting hepatic glucose production in relation to fasting glucose levels in healthy elderly controls The metabolic alterations which occur in middle-aged and elderly patients with diabetes. Hepatic glucose pro- subjects with type 2 diabetes have been extensively duction was measured by infusing radioactive glucose characterized [1]. When compared to age- and weight- tracers. matched controls, both lean and obese middle-aged 500 subjects have elevated fasting hepatic glucose Lean control production, a marked resistance to insulin-mediated Lean diabetes glucose disposal, and a profound impairment in 400 glucose-induced pancreatic insulin release. Recently, metabolic factors have been characterized 300 in lean and obese elderly patients with diabetes

[35–38]. These studies have demonstrated some Insulin (pM) 200 surprising differences in the metabolic profile between middle-aged and elderly subjects. In contrast to 100 younger subjects, fasting hepatic glucose production is normal in both lean and obese elderly subjects (Figure 1.2). Similar to younger subjects, lean elderly 500 patients have a profound impairment in pancreatic Obese control Obese diabetes insulin secretion but, in contrast to the young, these 400 patients have minimal resistance to insulin-mediated glucose disposal (Figures 1.3 and 1.4). In contra- 300 distinction to the young, obese elderly subjects have relatively preserved glucose-induced insulin Insulin (pM) 200 secretion (see Figure 1.3), although pulsatile insulin secretion is clearly altered [8]. Similar to the young however, these patients have a marked resistance to 100 insulin-mediated glucose disposal (Figure 1.4). In summary, the principal defect in lean elderly subjects 0 20 40 60 80 100 120 is impaired glucose-induced insulin release, while Time (min) the principal defect in obese patients is resistance to Figure 1.3 Glucose-induced insulin release in healthy insulin-mediated glucose disposal. elderly controls and elderly patients with diabetes. In- One of the most interesting findings of these studies sulin values were measured at glucose levels approxi- was that the ability of insulin to enhance blood flow mately 5 mmol l−1 above fasting levels. 6 CH 01 PATHOPHYSIOLOGY OF DIABETES IN THE ELDERLY

10 Middle-aged controls 6 Elderly controls 16% 8 Elderly diabetes ) 1 − 37% min 1 6 − ) 4 1 − min 1 − 4 (ml 100ml

Insulin-mediated blood flow 2 (mg kg 2

Insulin-mediated glucose disposal 0 60 120 180 0 Time (min) Lean Lean Obese Obese control diabetes control diabetes Figure 1.5 Insulin-mediated blood flow in obese middle-aged controls and obese elderly controls and Figure 1.4 Insulin-mediated glucose disposal rates patients with diabetes. Blood flow was measured in the in healthy elderly controls and elderly patients with calf during euglycaemic clamp studies utilizing venous diabetes. Glucose disposal rates were measured utilizing occlusion plethysmography. the euglycaemic clamp technique. In this technique, insulin is infused to achieve levels occurring after a glucose-induced insulin secretion seen in lean older meal, and glucose is infused simultaneously to prevent patients with type 2 diabetes. However, the clinical hypoglycaemia. significance of elevated antibodies in the elderly is was markedly reduced in obese, insulin-resistant older less certain. Some studies have found that elderly pa- patients (Figure 1.5) [37]. Insulin-mediated vasodi- tients with diabetes who are positive for GAD have lation is thought to account for about 30% of normal impaired beta-cell function relative to controls without glucose disposal, presumably because it increases the these antibodies, but others have not [41, 48]. It has delivery of insulin and glucose to muscle tissue. been suggested that screening for these auto-antibodies Indeed, it has been demonstrated that angiotensin- should be performed in elderly patients with impaired converting enzyme (ACE) inhibitors may improve glucose tolerance (IGT) and newly diagnosed diabetes, insulin sensitivity in elderly patients with diabetes in order to help predict which patients will develop and hypertension [39]. This suggests that drugs which islet cell failure. Although this is a compelling idea, we enhance muscle blood flow may prove to be valuable should only begin widespread screening when random- adjuncts in the future for the therapy of elderly ized studies have demonstrated that early intervention patients with diabetes. will protect the beta cells and reduce the need for in- It has been known for a number of years that autoim- sulin therapy [45–47]. Thus, it is unclear at present mune phenomena play a pivotal role in the beta-cell whether the measurement of auto-immune parameters failure that occurs in patients with type 1 diabetes can be used to predict future insulin requirements in [40]. It is also increasingly recognized that a subset the aged, or whether elderly patients with these abnor- of middle-aged patients with type 2 diabetes have a malities should be treated with therapies designed to form of diabetes that is characterized by beta-cell fail- modify autoimmune destruction of the pancreas. ure, and these patients often have high titres of islet Based on the above information, it is believed that cell antibodies and antibodies to glutamic acid de- the therapeutic approach to diabetes in the elderly carboxylase (GAD), similar to younger patients with should be different. In middle-aged patients, many en- type 1 diabetes. These patients have been said to docrinologists recommend that patients be treated with have LADA (latent autoimmune diabetes in adults) drugs that both stimulate insulin secretion and improve [41–47]. It is tempting to speculate that autoimmune insulin sensitivity, on the assumption that most pa- phenomena contribute to the profound impairment in tients have multiple metabolic problems. However, in 1.4 METABOLIC ALTERATIONS 7 lean elderly subjects, the principal defect is an impair- suggested that up to 80% of postprandial glucose ment in glucose-induced insulin secretion, and the main uptake in these patients may occur via glucose effec- approach should be to administer secretogogues to tiveness. At the present time it is uncertain whether stimulate insulin secretion, or to administer exogenous defects in NIMGU contribute to elevated glucose lev- insulin. In obese elderly patients, the principal defect is els in middle-aged patients with diabetes, as studies insulin resistance; hence, patients should be treated ini- which have evaluated this parameter in these patients tially with drugs that enhance insulin-mediated glucose have provided inconsistent results. disposal, such as metformin. It has been shown previously in healthy elderly subjects that glucose effectiveness is impaired during 1.4.1 The incretin pathway fasting, but is normal during hyperglycaemia [55]. Re- cently, it was demonstrated that elderly patients with The enteroinsular axis refers to hormones released diabetes have an even greater impairment in glucose from the gut in response to nutrient ingestion that result effectiveness than healthy elderly subjects (Figure 1.6) in enhanced glucose-induced insulin release, known as [56]. Although the cause of this abnormality is uncer- the ‘incretin effect’. The most important incretin hor- tain, it may relate to a decreased ability of glucose to mones are glucose-dependent insulinotropic polypep- recruit glucose transporters to the cell surface in these tide (GIP) and glucagon-like peptide 1 (GLP-1). patients. Although, both basal and glucose-stimulated GIP In the future, this metabolic abnormality may prove and GLP-1 levels have been found to be unchanged to be of great therapeutic relevance to the elderly. or to be increased in healthy elderly subjects when In younger patients, exercise, anabolic steroids and a compared to young controls, GIP and GLP-1 secretion reduction in free fatty acid levels have been shown is clearly enhanced in the elderly patient with diabetes to enhance glucose effectiveness [54]. Since we have [49, 50]. In addition, the levels of dipeptidyl peptidase shown that the incretin hormone GLP-1 may enhance IV (DPIV), the enzyme that breaks down GIP and NIMGU in elderly patients with diabetes [57], it is GLP-1, is progressively reduced with aging and possible that future therapies for the elderly may be diabetes. Beta-cell responses to GIP are reduced in directed not only at increasing insulin secretion and normal elderly subjects and are absent in elderly patients with diabetes [51, 52]. In contrast, beta-cell reversing insulin resistance, but also at enhancing glu- responses to GLP-1 are preserved in the elderly patient cose effectiveness. with diabetes [53]. These data suggest that GLP-1 and its analogues may prove to be useful therapeutic 0.015 p<0.0001 Control vs Diabetes options in the elderly, but that agents which prevent ) 1 the breakdown of GLP-1, such as DPIV inhibitors, −

may be less effective. min 1 − 0.01 1.4.2 Glucose effectiveness or non-insulin-mediated glucose uptake It has been recognized for many decades that insulin is an important hormone involved in the uptake of 0.005 glucose into cells. It has been demonstrated that glucose can stimulate its own uptake in the absence of insulin [54], an effect which is known as ‘glucose Glucose effectiveness (dl kg effectiveness’ or non-insulin-mediated glucose uptake 0 (NIMGU). Under fasting conditions, approximately Control Type 2 70% of glucose uptake occurs via glucose effective- diabetes ness, primarily in the central nervous system. After a Figure 1.6 Glucose effectiveness in elderly controls meal, approximately 50% of glucose uptake in normal and patients with diabetes. During these studies, insulin subjects occurs via NIMGU, with the bulk occurring secretion was suppressed by infusing the somatostatin in skeletal muscle. Because many middle-aged sub- analogue octreotide. Glucose was then infused to assess jects with diabetes are insulin-resistant, it has been glucose disposal in the absence of insulin. 8 CH 01 PATHOPHYSIOLOGY OF DIABETES IN THE ELDERLY 1.5 Molecular biology studies know how to interpret these symptoms when they occur [67]. At present, there is very little information available The most important hormone in the defence against regarding molecular biological abnormalities that may hypoglycaemia in normal subjects is glucagon. If be present in elderly patients with diabetes. The glu- glucagon responses are deficient, epinephrine becomes cokinase gene controls the glucose sensor for the beta important, and growth hormone and cortisol come cell, and defects in this gene could lead to the im- into play if hypoglycaemia is prolonged for several pairment in glucose-induced insulin secretion that is hours. The responses of both glucagon and growth present in lean elderly patients with diabetes. To date, hormone to hypoglycaemia are impaired in healthy elderly subjects, and to an even greater extent in evidence for mutations in this gene in elderly patients older patients with diabetes (Figure 1.7) [68]. Yet, is conflicting [58, 59]. even when they are educated about the symptoms of In skeletal muscle, insulin binds to its receptor, re- hypoglycaemia, the elderly have a reduced awareness sulting in activation of the insulin receptor tyrosine of the autonomic warning symptoms (sweating, kinase. Activation of this enzyme sets in motion a cas- palpitations, etc.) at glucose levels that would elicit a cade of intracellular events which although at present marked response in younger subjects. Consequently, is incompletely understood, results in the translocation the first symptoms of hypoglycaemia in the elderly of glucose transporters to the cell surface. In theory, a are often neuroglycopenic. Finally, elderly patients defect in any of these pathways could lead to insulin have an impaired psychomotor performance during resistance. To date, these intracellular processes have been incompletely studied in elderly patients with diabetes, but the preliminary information suggests that Glucose (mM) while insulin receptor numbers and affinity are nor- B 5.0 4.4 3.8 3.3 2.8 mal, the insulin receptor kinase activity may be de- Young fective [60]. Recent data have suggested that mito- 200

) Old 1 chondrial dysfunction contributes to insulin resistance − in middle-aged patients with diabetes, and potentially Diabetes also to impairments in glucose-induced insulin release 100 [61]. Although normal aging is characterized by progressive mitochondrial dysfunction, to date no studies Glucagon (ng l have been performed to assess mitochondrial function 0 in elderly patients with diabetes [62–64]. Clearly, further studies are required to elucidate the subcellular 30 defects that cause abnormal glucose metabolism in the elderly patient with diabetes. 20 ) 1 − g l 1.6 Glucose counter-regulation μ 10 GH( Numerous studies have demonstrated that elderly patients with diabetes, when compared to younger 0 patients, have an increased frequency of severe or fatal hypoglycaemia [12, 65, 66]. Recently, several 0 60 120 180 240 300 studies have evaluated glucose counter-regulation in Time (min) elderly subjects in an attempt to determine the cause Figure 1.7 Glucagon and growth hormone responses of the increased frequency of hypoglycaemia, and to hypoglycaemia in healthy young, healthy old and a number of important observations have emerged. elderly patients with diabetes. Controlled hypoglycaemia It appears that many elderly patients with diabetes was induced using the glucose clamp technique. Glucose have not been educated about the warning symptoms values at which hormone levels were measured are of hypoglycaemia and, as a consequence, do not shown on the top x-axis. 1.7 CONCLUSIONS 9 hypoglycaemia, which would prevent them from tak- and the Canadian Diabetes Association. I gratefully ing steps to return the blood glucose value to normal, acknowledge the support of the Allan McGavin Geri- even if they were aware that it was low. Thus, the atric Endowment at the University of British Columbia, increased frequency of hypoglycaemia in the elderly and the Jack Bell Geriatric Endowment Fund at Van- is due to a constellation of abnormalities, including couver Hospital and Health Science Centre. a reduced knowledge and awareness of the warning I am especially indebted to my longstanding col- symptoms, a decreased counter-regulatory hormone laborators in this work, particularly Dr Dariush Elahi secretion, and an altered psychomotor performance. at Johns Hopkins Medical School and Dr Daniel The levels of pancreatic polypeptide (PP) are Tessier at the University of Sherbrooke. I thank Rose- elevated during hypoglycaemia, and this response is marie Torressani, Eugene Mar, Gail Chin and Christine mediated by the vagus nerve. The role of PP in normal Lockhart for technical assistance in conducting these glucose counter-regulation is uncertain, but in younger studies. patients with diabetes a reduced PP response to hypoglycaemia is an early marker of autonomic insufﬁciency. Recently, it was found that although elderly patients with diabetes often have evidence References of autonomic dysfunction, their PP responses to 1. DeFronzo, R.A. (1988) Lilly Lecture 1987. The tri- hypoglycaemia are normal [69]. Thus, PP responses to umvirate: β-cell, muscle, liver. A collusion responsible hypoglycaemia cannot be used to predict autonomic for NIDDM. Diabetes, 37, 667–87. function in elderly patients. 2. Kahn, C.R. (1984) Banting Lecture. Insulin action, di- Based on the above information, there are a number abetogenes, and the cause of type II diabetes. Diabetes, of interventions that can be proposed to prevent hypo- 43, 1066–84. glycaemic events in the elderly. First, it would seem 3. Morris, R.D. and Rimm, A.A. (1991) Association of prudent to educate elderly patients about the warning waist to hip ratio and family history with the prevalence symptoms of hypoglycaemia so that they can appre- of NIDDM among 25,272 adult, white females. Am. J. ciate them when they occur. Second, consideration Public Health, 81, 507–9. should be given to the use of oral agents or insulin 4. Lipton, R.B., Liao, Y., Cao, G., Cooper, R.S. preparations that are associated with a lower frequency and McGee, D. (1993) Determinants of incident of hypoglycaemic events in the elderly. non-insulin-dependent diabetes mellitus among blacks and whites in a national sample. The NHANES 1 epidemiologic follow-up study. Am. J. Epidemiol., 1.7 Conclusions 138, 826–39. 5. Vaag, A., Henriksen, J.E., Madsbad, S., Holm, N. In summary, diabetes in the elderly is caused by a and Beck-Nielsen, H. (1995) Insulin secretion, insulin combination of genetic and environmental factors su- action, and hepatic glucose production in identical perimposed on the normal age-related changes in car- twins discordant for non-insulin-dependent diabetes bohydrate metabolism. The metabolic alterations that mellitus. J. Clin. Invest., 95, 690–8. 6. Muller, D.C. Elahi, D., Tobin, J.D. and Andres, R. occur in elderly patients with diabetes appear to be dis- (1996) The effect of age on insulin resistance and tinct from those that occur in younger patients. As we secretion: a review. Semin Nephrol., 16, 289–98. gain a greater appreciation of the pathophysiological 7. Iozzo, P., Beck-Nielsen, J., Laakso, M., Smith, U., abnormalities that occur in the elderly, we hope to be Yki-Jarvinen, H. and Ferrannini, E. (1999) Indepen- able to develop a more focused approach to therapy in dent inﬂuence of age on basal insulin secretion in this age group. It is only in this way that we will be non-diabetic humans. European group for the study of able to better cope with the epidemic of diabetes in the insulin resistance. Journal of Clinical Endocrinology elderly which will befall us in the coming decades. and Metabolism, 84, 863–8. 8. Meneilly, G.S., Veldhuis, J.D. and Elahi, D. (2005) De- convolution analysis of rapid insulin pulses before and Acknowledgments after six weeks of continuous subcutaneous administration of GLP-1 in elderly patients with type 2 diabetes. The studies described in this chapter were supported by Journal of Clinical Endocrinology and Metabolism, 90, grants from the Canadian Institutes of Health Research 6251–6. 10 CH 01 PATHOPHYSIOLOGY OF DIABETES IN THE ELDERLY

9. Zethelius, B., Lithell, H.O., Hales, C.N. and Berne, Alcohol consumption and risk of type 2 diabetes among C. (2004) Insulin resistance, impaired early insulin older women. Diabetes Care, 28, 2933–8. response, and insulin propeptides as predictors of the 23. Paolisso, G., D’Amore, A., Galzerano, D., Balbi, V., development of type 2 diabetes. Diabetes Care, 27, Giugliano, D., Varricchio, M. and D’Onofrio, F. (1993) 1433–8. Daily vitamin E supplements improve metabolic con- 10. Ferrannini, E. and the European Group for the Study trol but not insulin secretion in elderly type II diabetic of Insulin Resistance. (1996) Insulin action and age. patients. Diabetes Care, 16, 1433–7. Diabetes, 45, 949. 24. Paolisso, G., D’Amore, A., Balbi, V., Volpe, C., Galz- 11. Pandit, M.K., Burke, J., Gustafson, A.B., Minocha, erano, D., Giugliano, D., Sgambato, S., Varricchio, A. and Peiris, A.N. (1993) Drug-induced disorders of M. and D’Onofrio, F. (1994) Plasma vitamin C af- glucose tolerance. Ann. Intern. Med., 118, 529–39. fects glucose homeostasis in healthy subjects and in 12. Meneilly, G.S. and Tessier, D. (2001) Diabetes in non-insulin-dependent diabetics. Am. J. Physiol., 266, elderly adults. J. Gerontol., 56A, M5–13. E261–8. 13. van Dam, R.M., Rimm, E.B., Willett, W.C. et al. (2002) 25. Paolisso, G., Scheen, A., Cozzolino, D. et al. (1994) Dietary patterns and risk for Type 2 diabetes mellitus Changes in glucose turnover parameters and improve- in U.S. men. Ann. Intern. Med., 136, 201–9. ment of glucose oxidation after 4-week magnesium 14. The DECODE-DECODA Study Group, on behalf of administration in elderly non-insulin-dependent (type the European Diabetes Epidemiology Group and the In- II) diabetic patients. Journal of Clinical Endocrinology ternational Diabetes Epidemiology Group. (2003) Age, and Metabolism, 78, 1510–15. body mass index and Type 2 diabetes - associations 26. Song, M.K., Rosenthal, M.J., Naliboff, B.D., modified by ethnicity. Diabetologia, 48, 1063–70. Phanumas, L. and Kang, K.W. (1998) Effects of 15. Goodpaster, B.H., Krishnaswami, S., Resnick, H. et al. bovine prostate powder on zinc, glucose and insulin metabolism in old patients with non-insulin-dependent (2003) Association between regional adipose tissue diabetes mellitus. Metabolism, 47, 39–43. distribution and both Type 2 diabetes and impaired 27. Lee, D.H., Folsom, A.R. and Jacobs, D.R. (2004) glucose tolerance in elderly men and women. Diabetes Dietary iron intake and Type 2 diabetes incidence in Care, 26, 372–9. postmenopausal women: the Iowa Women’s Health 16. Meigs, J.B., Muller, D.C., Nathan, D.M. et al. (2003) Study. Diabetologia, 47, 185–94. The natural history of progression from normal glucose 28. Barzilay, J.I., Abraham, L., Heckbert, S. et al. (2001) tolerance to Type 2 diabetes in the Baltimore longitu- The relation of markers of inflammation to the devel- dinal study of aging. Diabetes, 52, 1475–84. opment of glucose disorders in the elderly. Diabetes, 17. Cassano, P.A., Rosner, B., Vokonas, P.S. and Weiss, 50, 2384–9. S.T. (1992) Obesity and body fat distribution in re- 29. Kanaya, A.M., Harris, T., Goodpaster, B.H. et al. lation to the incidence of non-insulin-dependent dia- (2004) Adipocytokines attenuate the association be- betes mellitus. American Journal of Epidemiology, 136, tween visceral adiposity and diabetes in older adults. 1474–86. Diabetes Care, 27, 1375–80. 18. Diabetes Prevention Program Research Group. (2002) 30. Lechleitner, M., Herold, M., Dzien-Bischinger, C., Reduction in the incidence of Type 2 diabetes with Hoppichlert, F. and Dzien, A. (2002) Tumour necro- lifestyle intervention or metformin. N. Engl. J. Med., sis factor-alpha plasma levels in elderly patients with 346, 393–403. Type 2 diabetes mellitus - observations over 2 years. 19. Feskens, E.J.M., Virtanen, S.M., Rasanen, L., Tuomile- Diabetic Medicine, 19, 949–53. hto, J., Stengard, J., Pekkanen, J., Nissinen, A. and 31. de Rekeneire, N., Peila, R., Ding, J., Kritchevsky, S.B., Kromhout, D. (1995) Dietary factors determining di- Colbert, L.H., Visser, M., Shorr, R.I., Kuller, L.H., abetes and impaired glucose tolerance. Diabetes Care, Strotmeyer, E.S., Schwartz, A.V., Vellas, B. and Harris, 18, 1104–12. T.B. (2006) Diabetes, hyperglycemia, and inflamma- 20. Song, Y., Manson, J.E., Buring, J.E. et al. (2004) A tion in older individuals. Diabetes Care, 29, 1902–8. prospective study of red meat consumption and Type 2 32. Snijder, M.B., Seidell, J.C., Heine, R.J., Bouter, L.M., diabetes in middle-aged and elderly women. Diabetes Nijpels, G., Stehouwer, C.D.A., Funahashi, T., Mat- Care, 27, 2108–15. suzawa, Y., Shimomura, I. and Dekker, J.M. (2006) 21. Meyer, K.A., Kushi, L.H., Jacobs, D.R. et al. (2000) Associations of adiponectin levels with incident im- Carbohydrates, dietary fiber, and incident type 2 dia- paired glucose metabolism and type 2 diabetes in older betes in older women. Am. J. Clin. Nutr., 71, 921–30. men and women. Diabetes Care, 29, 2498–503. 22. Beulens, J.W.J., Stolk, R.P., Van der Schouw, Y.T., 33. Oh, J.-Y., Barrett-Connor, E., Wedick, N.M. et al. Grobbee, D.E., Hendriks, H.F.J. and Bots, M.L. (2005) (2002) Endogenous sex hormones and the development 1.7 CONCLUSIONS 11

of Type 2 diabetes in older men and women: the 47. Leslie, R.D.G., Williams, R. and Pozzilli, P. (2006) Rancho Bernardo Study. Diabetes Care, 25, 55–60. Clinical review: type 1 diabetes and latent autoimmune 34. Golden, S.H., Dobs, A.S., Vaidya, D., Szklo, M., Gap- diabetes in adults: One end of the rainbow. Journal of stur, S., Kopp, P., Liu, K, and Ouyang, P. (2007) Clinical Endocrinology and Metabolism, 91, 1654–9. Endogenous sex hormones and glucose tolerance sta- 48. Meneilly, G.S., Tildesley, H., Elliott, T., Palmer, J.P. tus in postmenopausal women. Journal of Clinical En- and Juneja, R. (2000) Signiﬁcance of GAD positivity in docrinology and Metabolism, 92 (4), 1289–95. elderly patients with diabetes. Diabetic Medicine, 17, 35. Arner, P., Pollare, T. and Lithell, H. (1991) Different 247–8. aetiologies of type 2 (non-insulin-dependent) diabetes 49. Meneilly, G.S., Demuth, H.-U., McIntosh, C.H.S. and mellitus in obese and non-obese subjects. Diabetologia, Pederson, R.A. (2000) Effect of ageing and diabetes 34, 483–7. on glucose-dependent insulinotropic polypeptide and 36. Meneilly, G.S., Hards, L., Tessier, D., Elliott, T. and dipeptidyl peptidase IV responses to oral glucose. Di- Tildesley, H. (1996) NIDDM in the elderly. Diabetes abetic Medicine, 17, 346–50. Care, 19, 1320–75. 50. Korosi, J., McIntosh, C.H.S., Pederson, R.A., Demuth, 37. Meneilly, G.S. and Elliott, T. (1999) Metabolic alter- H.-U., Habener, J.F., Gingerich, R., Egan, J.M., Elahi, ations in middle-aged and elderly obese patients with D. and Meneilly, G.S. (2001) Effect of aging and dia- type 2 diabetes. Diabetes Care, 22, 112–18. betes on the enteroinsular axis. Journal of Gerontology, 38. Meneilly, G.S. and Elahi, D. (2005) Metabolic alter- 56A, M575–9. ations in middle-aged and elderly lean patients with 51. Meneilly, G.S., Ryan, A.S., Minaker, K.L. and Elahi, Type 2 diabetes. Diabetes Care, 28, 1498–9. D. (1998) The effect of age and glycemic level on 39. Paolisso, G., Gambardella, A., Verza, M., D’Amora, the response of the α-cell to glucose-dependent in- A., Sgambato, S. and Varricchio, M. (1992) ACE- sulinotropic polypeptide and peripheral tissue sensitivity to endogenously released insulin. Journal of Clinical inhibition improves insulin-sensitivity in age insulin- Endocrinology and Metabolism, 83, 2925–31. resistant hypertensive patients. J. Hum. Hypertens., 6, 52. Elahi, D., McAloon-Dyke, M., Fukagawa, N.K., 175–9. Meneilly, G.S., Sclater, A.L., Minaker, K.L., Habener, 40. Zimmet, P.Z. (1999) Diabetes epidemiology as a tool J.F. and Andersen, D.K. (1994) The insulinotropic ac- to trigger diabetes research and care. Diabetologia, 42, tions of glucose-dependent insulinotropic polypeptide 499–518. (GIP) and glucagon-like peptide-1 (7-37) in normal 41. Monge, L., Brunot, G., Pinach, S., et al. (2004) A and diabetic subjects. Regul. Pept., 51, 63–74. clinically orientated approach increases the efﬁciency 53. Meneilly, G.S., McIntosh, C.H., Pederson, R.A., of screening for latent autoimmune diabetes in adults Habener, J.F., Gingerich, R., Egan, J.M. and Elahi, (LADA) in a large clinic-based cohort of patients with D. (2001) Glucagon-like peptide-1 (7-37) augments diabetes onset over 50 years. Diabetic Medicine, 21, insulin release in elderly patients with diabetes. 456–9. Diabetes Care, 24, 964–5. 42. Barinas-Mitchell, E., Pietropaolo, S., Zhang, Y.-J. et al. 54. Best, J.D., Kahn, S.E., Ader, M., Watanabe, R.M., (2004) Islet cell autoimmunity in a triethnic adult Ni, T.C. and Bergman, R.N. (1996) Role of glucose population of the Third National Health and Nutrition effectiveness in the determination of glucose tolerance. Examination Survey. Diabetes, 53, 1293–302. Diabetes Care, 19, 1018–30. 43. Zinman, B., Kahn, S.E., Haffner, S.M. et al. (2004) 55. Meneilly, G.S., Elahi, D., Minaker, K.L., Sclater, Phenotypic characteristics of GAD antibody-positive A.L. and Rowe, J.W. (1989) Impairment of recently diagnosed patients with Type 2 diabetes in noninsulin-mediated glucose disposal in the elderly. North America and Europe. Diabetes, 53, 3193–200. Journal of Clinical Endocrinology and Metabolism, 44. Pozzilli, P. and Di Mario, U. (2001) Autoimmune di- 63, 566–71. abetes not requiring insulin at diagnosis (Latent au- 56. Forbes, A., Elliott, T., Tildesley, H., Finegood, D. toimmune diabetes of the adult). Diabetes Care, 24, and Meneilly, G.S. (1998) Alterations in non-insulin- 1460–7. mediated glucose uptake in the elderly patient with 45. Gale, E.A.M. (2005) Latent autoimmune diabetes diabetes. Diabetes, 47, 1915–19. in adults: a guide for perplexed. Diabetologia, 48, 57. Meneilly, G.S., McIntosh, C.H., Pederson, R.A., 2195–9. Habener, J.F., Gingerich, R., Egan, J.M., Finegood, 46. Fourlanos, S., Dotta, F., Greenbaum, C.J., Palmer, D.T. and Elahi, D. (2001) Effect of glucagon-like J.P., Rolandsson, O., Colman, P.G. and Harrison, L.C. peptide 1 on non-insulin-mediated glucose uptake in (2005) Latent autoimmune diabetes in adults (LADA) the elderly patient with diabetes. Diabetes Care, 24, should be less latent. Diabetologia, 48, 2206–12. 1951–6. 12 CH 01 PATHOPHYSIOLOGY OF DIABETES IN THE ELDERLY

58. Laakso, M., Malkki, M., Kekalainen, P., Kuusisto, J., the elderly: possible role in insulin resistance. Science, Mykkanen, L. and Deeb, S.S. (1995) Glucokinase gene 300, 1140–2. variants in subjects with late-onset NIDDM and im- 64. Ritz, P. and Berrut, G. (2005) Mitochondrial function, paired glucose tolerance. Diabetes Care, 18, 398–400. energy expenditure, aging and insulin resistance. Dia- 59. McCarthy, M.I., Hitman, G.A., Hitchins, M. et al. betes Metab., 31, 5S67–5S73. (1993) Glucokinase gene polymorphisms; a genetic 65. Stepka, M., Rogala, H. and Czyzyk, A. (1993) Hy- marker for glucose intolerance in cohort of elderly poglycemia; a major problem in the management of Finnish men. Diabetic Med., 10, 198–204. diabetes in the elderly. Aging, 5, 117–21. 60. Obermajer-Kusser, B., White, M.F., Pongratz, D.E. 66. Lassmann-Vague, L. (2005) Hypoglycemia in elderly et al. (1989) A defective intramolecular autoactiva- diabetic patients. Diabetes Metab., 31, 5S53–5S57. tion cascade may cause the reduced kinase activity 67. Thomson, F.J., Masson, E.A., Leeming, J.T. and Boul- of the skeletal muscle insulin receptor from patients ton, A.J. (1991) Lack of knowledge of symptoms of with non-insulin-dependent diabetes mellitus. J. Biol. hypoglycaemia by elderly diabetic patients. Age and Chem., 264, 9497–504. Aging, 20, 404–6. 61. Hawley, J.A. and Lessard, S.J. (2007) Mitochondrial 68. Meneilly, G.S., Cheung, E. and Tuokko, H. (1994) function: use it or lose it. Diabetologia, 50, 699–702. Counterregulatory hormone responses to hypoglycemia 62. Barazzoni, R. (2004) Skeletal muscle mitochondrial in the elderly patient with diabetes. Diabetes, 43, protein metabolism and function in ageing and type 403–10. 2 diabetes. Current Opinion in Clinical Nutrition and 69. Meneilly, G.S. (1996) Pancreatic polypeptide responses Metabolic Care, 7, 97–102. to hypoglycemia in aging and diabetes. Diabetes Care, 63. Petersen, K.F., Befroy, D., Dufour, S., Dziura, J., 19, 544–6. Ariyan, C., Rothman, D.L., DiPietro, L., Cline, G.W. and Shulman, G.I. (2003) Mitochondrial dysfunction in 2 Diabetes-Related Risk Factors In Older People

Stephen Colagiuri and Daniel Davies The University of Sydney, Institute of Obesity Nutrition and Exercise, Medical Foundation Building, NSW, Australia

we review the established risk factors, with particular Key messages emphasis on older people. • Older people with undiagnosed type 2 diabetes often have easily identifiable risk factors for this 2.2 Age condition. • The presence of cardiovascular or cerebrovas- Populations worldwide continue to show a consistent cular disease in an older individual should be a increase in the prevalence of diagnosed and undi- prompt for the detection of diabetes. agnosed type 2 diabetes with increasing age, with • Specific risk factor studies in older people with values reaching a plateau or even declining slightly type 2 diabetes are lacking; this emphasizes the in the very old. For example, in Australia among importance of further research in this area. the 25–34 year age group, 0.2% of the subjects have diagnosed and 0.1% undiagnosed diabetes, these values increasing respectively to 9.4% and 8.5% 2.1 Introduction among 65–74 year olds and to 10.9% and 12.1% for those aged 75 years and more [1]. In the US, among There are a number of well-established and emerging the age groups of 70–74, 75–79, 80–84 and ≥85 risk factors which are associated with undiagnosed type years, the prevalence of diabetes was shown to be 2 diabetes, and which predict the future development 20.0, 21.1, 20.2 and 17.3%, respectively [2]. of this condition. Most people with undiagnosed type 2 In the DECODE study, the data analyzed from nine diabetes have easily identifiable risk factors, and these European countries indicated a prevalence of type 2 form the basis for the targeted testing of high-risk diabetes of <10% among people aged <60 years, and groups, not only to diagnose the condition but also to of 10–20% in those aged 60–79 years [3]. The findings facilitate its earlier treatment. Similar risk factors can of the DECODA study in 11 Asian cohorts were be used to screen for individuals at high risk of the similar, with the overall prevalence of diabetes (both future development of diabetes, who can be targeted known and undiagnosed) increasing with age, peaking for diabetes-prevention programs. at 70–89 years in Chinese and Japanese subjects and at While few studies have specifically examined the 60–69 years, followed by a decline beyond 70 years, risk factors for diabetes in the elderly, in this chapter in Indian subjects [4].

Diabetes in Old Age Third Edition Edited by Alan J. Sinclair © 2009 John Wiley & Sons Ltd. ISBN: 978-0-470-06562-4 14 CH 02 DIABETES-RELATED RISK FACTORS IN OLDER PEOPLE 2.3 Impaired glucose tolerance 2.5 Ethnicity (IGT) and impaired fasting glucose (IFG) It is well documented that the prevalence of diabetes varies among different ethnic groups [11], and this difference is also evident in older people living in the The prevalence of IGT and IFG increases with age. For same country. McBean and colleagues [2] analyzed example, in the NHANES III study [5], the prevalence a random sample of US Medicare beneficiaries to of IGT increased from 11.1% in people aged 40–49 examine the prevalence of diabetes in people aged ≥67 years to 20.9% in those aged 60–74 years. years from different racial/ethnic groups. Although Both IGT and IFG are important risk factors for the prevalence of diabetes was high in all groups, the development of future diabetes, increasing the risk it was significantly higher among Hispanics (33.4%) between 10- and 20-fold compared to those subjects than among blacks (29.6%), Asians (24.3%) and whites with normal glucose tolerance [6]. This increased risk (18.4%, p < 0.0001). does not seem to vary with age, however. In the US Diabetes Prevention Program, the incidence of type 2.6 Gestational diabetes mellitus 2 diabetes was 11.0 cases per 100 person-years in the placebo group overall, compared to 10.8 cases (GDM) per 100 person-years in those aged 60 and above Many studies have reported that women with a [7]. previous history of GDM are at increased risk of Several studies have shown that progression to di- developing type 2 diabetes. abetes can be either prevented or delayed in people In a systematic review of 28 studies which examined with IGT [7, 8]. In the US Diabetes Prevention Pro- the incidence of type 2 diabetes following GDM [12], gram, lifestyle modification achieved a 58% reduction the cumulative incidence ranged from 2.6% to 70%, in terms of diabetes progression, compared to a 31% with the follow-up period ranging from 6 weeks to 28 reduction achieved with metformin treatment. The ef- years postpartum. The cumulative incidence increased fect of lifestyle modification was greatest in people substantially during the first 5 years after delivery, but aged ≥60 years, whereas the effect of metformin was then increased more slowly after 10 years. However, not significant in this age group [7]. as so few studies have been reported with long-term follow up, it is difficult to draw firm conclusions regarding the influence of previous GDM on diabetes 2.4 Body weight risk during old age.

An increasing body weight is associated with undiagnosed diabetes, and is also a risk factor for future 2.7 Family history diabetes. A body mass index (BMI) of ≥30 kg m−2 increases the absolute risk of type 2 diabetes by ap- Individuals with a family history of type 2 diabetes are at an increased risk for the disease. The lifetime risk of proximately twofold, but up to 20-fold, relative to a developing type 2 diabetes is estimated at 40% if one normal BMI. parent has type 2 diabetes [13]. Although most studies The association between obesity and type 2 diabetes have reported that the effect is not gender-specific, is also observed in older people from different regions there have been some exceptions. For example, Mooy ≥ of the world. In a French study of people aged 60 et al. [14] reported a positive effect only in males, and years, the prevalence of type 2 diabetes was threefold Sugimori and coworkers [15] only in females. higher in the upper quartile of BMI compared to the The effect of family history is seen across age lowest quartile in both men and women [9]. In a groups. For example, when Costa et al. [16] studied Taiwanese study of people with a mean age of 72.8 205 non-diabetic siblings of people with type 2 years, a 2.5-fold increased risk of type 2 diabetes was diabetes, compared to the general population – and observed with increasing BMI [10]. within any age group – type 2 diabetes was more 2.11 ANTIPSYCHOTIC MEDICATION AND MENTAL ILLNESS 15 common among those people with a family history of abnormalities in 238 people (mean age ca. 70 years) the condition. with acute stroke, they found 16% of the patients to have newly diagnosed type 2 diabetes, based on an OGTT. The prevalence of diabetes at discharge in 2.8 Hypertension 106 people (median age 71 years) with acute ischemic stroke and no history of known diabetes was 46%, A number of studies have shown that hypertension also based on an OGTT [27]. On admission, however, is associated with an approximate twofold increase a further 29% already had a diagnosis of diabetes. in undiagnosed type 2 diabetes [17–21]. This increase appears to be uniform across age groups; for example, when Bog-Hansen et al. [22] examined a 2.10 Physical inactivity community-based population with hypertension for the presence of undiagnosed type 2 diabetes, the over- There is a clear relationship between (a lack of) physi- all prevalence of the previously undiagnosed condition cal activity and the development of type 2 diabetes. A ranged from 17% to 26% in those aged <70 years, and meta-analysis of 10 prospective cohort studies which ≥ 31% in people aged 70 years. included a total of 301 221 participants showed that, compared to being sedentary, the participation in mod- 2.9 Cardiovascular and erate intensity physical activity had a relative risk (RR) cerebrovascular disease of 0.69 for developing type 2 diabetes. In five of these cohort studies that specifically investigated the role of walking, the RR of type 2 diabetes was 0.70 for reg- Diabetes, whether diagnosed or undiagnosed, is com- ≥ mon in people with cardiovascular and cerebrovas- ular walking (usually 2.5 h per week brisk walking) cular disease. Consequently, testing for undiagnosed compared with almost no walking [31]. diabetes is routinely recommended in all people with Less information is available on the relationship be- macrovascular disease, including the elderly. tween physical activity and undiagnosed diabetes. Baan A number of studies have reported a high preva- and colleagues [32] assessed this relationship in 1016 lence of undiagnosed diabetes among individuals with people aged 55–75 years and without known diabetes, an acute myocardial infarction. Among 4961 subjects from the Rotterdam Study. The total time spent on (median age 66 years) from 25 European countries, physical activity per week was seen to decrease with all of whom had coronary artery disease (CAD), type increasing glucose intolerance, with adjusted odds ra- 2 diabetes had been previously diagnosed in 29% of tios (ORs) for vigorous activities such as cycling (0.26 cases. Among 1920 subjects without known diabetes in men, 0.37 in women) and other sports (0.28 in men) who underwent an oral glucose tolerance test (OGTT), showing an inverse association with the prevalence of newly diagnosed diabetes was found to be present in newly diagnosed diabetes. 22% of those with acute CAD, and in 14% of those In an elderly French population (n = 2532, all aged with stable CAD [23]. ≥60 years), sporting activity showed a negative inde- Among subjects aged ≥80 years with acute my- pendent association with the prevalence of diagnosed ocardial infarction in the Glucose tolerance in Acute and undiagnosed type 2 diabetes, with a significantly Myocardial Infarction (GAMI) study, the prevalence lower prevalence of type 2 diabetes in those who of type 2 diabetes at discharge was 34% [24]. Of these played for at least 30 min per day, compared to those patients, 93% still had abnormalities of glucose toler- who played for less than 30 min (OR 0.61 in men, 0.62 ance after 12 months (64% with type 2 diabetes, 29% in women) [9]. with IGT). In a cohort of 3266 people (mean age ca. 65 years) scheduled for coronary angiography, the prevalence of 2.11 Antipsychotic medication diabetes was 32% (17% known diabetes, 15% undiag- and Mental Illness nosed diabetes) [25]. Diabetes is also common in patients with cere- Certain mental illnesses and antipsychotic medications brovascular disease. For example, when Matz and appear to be associated with an increase in type 2 colleagues [26] examined the prevalence of glucose diabetes, although study findings vary widely. 16 CH 02 DIABETES-RELATED RISK FACTORS IN OLDER PEOPLE In a retrospective, chart-review in the US, the preva- the increased risk is due to the schizophrenia itself or lence of type 2 diabetes was assessed in 243 psychiatric to the antipsychotic treatment. inpatients aged 50–74 years with a variety of mental illnesses [33]. The overall prevalence of type 2 diabetes was 25%, which was significantly greater than the rate 2.12 Sleep disorders expected for an age-, race- and gender-matched group in the general US population (14%). The rates of type The association of sleep-disordered breathing and type 2 diabetes for each mental illness were: schizoaffective 2 diabetes is increasingly recognized [37]. Among disorder (50%)bipolar I disorder (26%)major depres- 2656 subjects (median age 68 years) from the Sleep sion (18%) = dementia (18%)schizophrenia (13%) (p Heart Health Study [38], the prevalence of diabetes <0.006), independent of the effects of age, race, gen- was increased approximately twofold in people with der, medication and body mass. Of these rates of type an elevated respiratory disturbance index. Hence, those 2 diabetes, only those for schizoaffective disorder and with this condition should be assessed routinely for bipolar I disorder were significantly higher than the undiagnosed diabetes. national norms. The data for antipsychotic medications were less 2.13 Smoking clear. Here, a meta-analysis of 25 observational pharmaco-epidemiological studies found no significant A meta-analysis of 25 prospective cohort studies in- difference in the risk of developing treatment-requiring volving 1.2 million participants found that active smok- type 2 diabetes using either second- or first-generation ing is associated with an increased risk of type 2 antipsychotics [34]. However, the data on two recent diabetes [39]. The pooled adjusted RR of type 2 di- second-generation antipsychotics, aripiprazole and abetes for active smoking compared to non-smoking ziprasidone, were limited. was 1.44. Moreover, the risk of type 2 diabetes was A systematic review of 17 pharmaco-epidemiologic greater for heavy smokers (≥20 cigarettes per day; studies examined the relationship between certain RR = 1.61) than for lighter smokers (RR = 1.29), and atypical antipsychotics and the risk of type 2 diabetes lower for former smokers (RR = 1.23) compared to [35]. Treatment with olanzapine in people with major active smokers (RR = 1.44). psychiatric illness, compared to no treatment, was associated with a significantly greater risk of new-onset diabetes. Risperidone was not associated with a greater 2.14 Conclusions RR of diabetes than conventional antipsychotics or no treatment. Of nine studies that compared the Diabetes-related risk factors are relevant and impor- RR of diabetes with risperidone and olanzapine, tant in older people, and should be strongly considered six demonstrated a significantly greater risk with when assessing the elderly for undiagnosed diabetes, olanzapine, although the magnitude of the risk varied and for the future risk of developing the condition. considerably across studies. Definitive conclusions While undiagnosed diabetes carries significant morbid- could not be drawn for clozapine and quetiapine, ity in the elderly, diabetes prevention studies generally due to insufficient evidence. Results from the review demonstrate equal or greater effectiveness when ap- also showed that three out of four studies did not plied to older rather than to younger people. demonstrate any significant increase in risk for diabetes using atypical antipsychotics compared to conventional antipsychotics. References Bellantuono and colleagues [36] reviewed 21 1. Dunstan DW, Zimmet PZ, Welborn TA, De Courten studies (nine prospective, 11 retrospective) to evaluate MP, Cameron AJ, Sicree RA, Dwyer T, Colagiuri the risk of type 2 diabetes in people treated with S, Jolley D, Knuiman M, Atkins R and Shaw JE. different antipsychotic drugs (conventional and (2002) The rising prevalence of diabetes and impaired second-generation). Subjects with schizophrenia glucose tolerance: the Australian Diabetes, Obesity and treated with different antipsychotics had a higher Lifestyle Study. Diabetes Care, 25 (5), 829–34. risk of developing type 2 diabetes than the general 2. McBean AM, Li S, Gilbertson DT and Collins AJ. population. It is not currently clear, however, whether (2004) Differences in diabetes prevalence, incidence, 2.14 CONCLUSIONS 17

and mortality among the elderly of four racial/ethnic 12. Kim C, Newton KM and Knopp RH. (2002) Gesta- groups: whites, blacks, Hispanics, and Asians. Diabetes tional diabetes and the incidence of type 2 diabetes: a Care, 27 (10), 2317–24. systematic review. Diabetes Care, 25 (10), 1862–8. 3. DECODE Study Group. (2003) Age- and sex-specific 13. Kobberling J and Tillil H. (1982) Empirical risk figures prevalences of diabetes and impaired glucose regula- for first degree relatives of non-insulin dependent dia- tion in 13 European cohorts. Diabetes Care, 26 (1), betics. In: The Genetics of Diabetes Mellitus. J. Kob- 61–9. berling and R. Tattershall (eds). Academic Press, Lon- 4. Qiao Q, Hu G, Tuomilehto J, Nakagami T, Balkau B, don, New York, pp. 201–9. Borch-Johnsen K, Ramachandran A, Mohan V, Iyer 14. Mooy JM, Grootenhuis PA, de Vires H, Valkenburg SR, Tominaga M, Kiyohara Y, Kato I, Okubo K, Nagai HA, Bouter LM, Kostense PJ and Heine RJ. (1995) M, Shibazaki S, Yang Z, Tong Z, Fan Q, Wang B, Chew Prevalence and determinants of glucose intolerance SK,TanBY,HengD,EmmanuelS,TajimaN,Iwamoto in a Dutch Caucasian population. The Hoorn study. Y, Snehalatha C, Vijay V, Kapur A, Dong Y, Nan H, Diabetes Care, 18, 1270–3. Gao W, Shi H and Fu F. (2003) Age- and sex-specific 15. Sugimori H, Miyakawa M, Yoshida K, Izuno T, Taka- prevalence of diabetes and impaired glucose regulation hashi E, Tanaka C, Nakamura K and Hinohara S. in 11 Asian cohorts. Diabetes Care, 26 (6), 1770–80. (1998) Health risk assessment for diabetes mellitus 5. Harris MI, Flegal KM, Cowie CC, Eberhardt MS, based on longitudinal analysis of MHTS database. Goldstein DE, Little RR, Wiedmeyer HM and J Med Systems, 22, 27–32. Byrd-Holt DD. (1998) Prevalence of diabetes, 16. Costa A, Rios M, Casamitjana R, Gomis R and Con- impaired fasting glucose, and impaired glucose get I. (1998) High prevalence of abnormal glucose tolerance in U.S. adults. The Third National Health tolerance and metabolic disturbances in first degree and Nutrition Examination Survey, 1988-1994. relatives of NIDDM patients. A study in Catalonia, a Diabetes Care, 21 (4), 518–24. Mediterranean community. Diabetes Res Clin Pract, 41, 6. Magliano DJ, Barr EL, Zimmet PZ, Cameron AJ, Dun- 191–6. stan DW, Colagiuri S, Jolley D, Owen N, Phillips 17. Saad MF, Knowler WC, Pettitt DJ, Nelson RG, Mott P, Tapp RJ, Welborn TA and Shaw JE. (2008) Glu- DM and Bennett PH. (1990) Insulin and hypertension. cose indices, health behaviors, and incidence of dia- Relationship to obesity and glucose intolerance in Pima betes in Australia: the Australian Diabetes, Obesity and Indians. Diabetes, 39, 1430–5. Lifestyle Study. Diabetes Care, 31 (2), 267–72. 18. Chou P, Liao MJ and Tsai ST. (1994) Associated risk 7. Knowler WC, Barrett-Connor E, Fowler SE, Hamman factors of diabetes in Kin-Hu, Kinmen. Diabetes Res RF, Lachin JM, Walker EA and Nathan DM. (2002) Clin Pract, 26, 229–35. Reduction in the incidence of type 2 diabetes with 19. Ruige JB, de Neeling JN, Kostense PJ, Bouter LM and lifestyle intervention or metformin. N Engl J Med, 346 Heine RJ. (1997) Performance of an NIDDM screening (6), 393–403. questionnaire based on symptoms and risk factors. 8. Tuomilehto J, Lindstrom J, Eriksson JG, Valle Diabetes Care, 20, 491–6. TT, Hamalainen H, Ilanne-Parikka P, Keinanen- 20. Welborn TA, Reid CM and Marriott G. (1997) Aus- Kiukaanniemi S, Laakso M, Louheranta A, Rastas M, tralian diabetes screening study: impaired glucose tol- Salminen V and Uusitupa M. (2001) Prevention of erance and non-insulin-dependent diabetes mellitus. type 2 diabetes mellitus by changes in lifestyle among Metabolism, 46 (Suppl. 1), 1–5. subjects with impaired glucose tolerance. N Engl 21. Baan CA, Ruige JB, Stolk RP, Witteman JCM, Dekker J Med, 344 (18), 1343–50. JM, Heine RJ and Feskens EJM. (1999) Performance 9. Defay R, Delcourt C, Ranvier M, Lacroux A and of a predictive model to identify undiagnosed diabetes Papoz L. (2001) Relationships between physical ac- in a health care setting. Diabetes Care, 22, 213–19. tivity, obesity and diabetes mellitus in a French elderly 22. Bog-Hansen E, Lindblad U, Bengtsson K, Ranstam population: the POLA study. Pathologies Oculaires J, Melander A and Rastam L. (1998) Risk fac- liees a l’Age. Int J Obes Relat Metab Disord, 25 (4), tor clustering in patients with hypertension and 512–18. non-insulin-dependent diabetes mellitus. The 10. Huang KC, Lee MS, Lee SD, Chang YH, Lin YC, Tu Skaraborg Hypertension Project. J Intern Med, 243, SH and Pan WH. (2005) Obesity in the elderly and its 223–32. relationship with cardiovascular risk factors in Taiwan. 23. Bartnik M, Ryden L, Ferrari R, Malmberg K, Pyorala Obes Res, 13 (1), 170–8. K, Simoons M, Standl E, Soler-Soler J and Ohrvik J. 11. International Diabetes Federation. (2006) Diabetes (2004) The prevalence of abnormal glucose regulation Atlas, Third Edition. in patients with coronary artery disease across Europe. 18 CH 02 DIABETES-RELATED RISK FACTORS IN OLDER PEOPLE

The Euro Heart Survey on diabetes and the heart. Eur 31. Jeon CY, Lokken RP, Hu FB and Van Dam RM. (2007) Heart J, 25 (21), 1880–90. Physical activity of moderate intensity and risk of type 24. Wallander M, Malmberg K, Norhammar A, Ryden L 2 diabetes: A systematic review. Diabetes Care, 30 (3), and Tenerz A. (2008) Oral glucose tolerance test: a re- 744–52. liable tool for early detection of glucose abnormalities 32. Baan CA, Stolk RP, Grobbee DE, Witteman JC and in patients with acute myocardial infarction in clinical Feskens EJ. (1999) Physical activity in elderly subjects practice: a report on repeated oral glucose tolerance with impaired glucose tolerance and newly diagnosed tests from the GAMI study. Diabetes Care, 31 (1), diabetes mellitus. American Journal of Epidemiology, 36–8. 149 (3), 219–27. 25. Taubert G, Winkelmann BR, Schleiffer T, Marz W, 33. Regenold WT, Thapar RK, Marano C, Gavirneni S Winkler R, Gok R, Klein B, Schneider S and Boehm and Kondapavuluru PV. (2002) Increased prevalence BO. (2003) Prevalence, predictors, and consequences of type 2 diabetes mellitus among psychiatric inpa- of unrecognized diabetes mellitus in 3266 patients tients with bipolar I affective and schizoaffective disor- scheduled for coronary angiography. Am Heart J, 145 ders independent of psychotropic drug use. Journal of (2), 285–91. Affective Disorders, 70 (1), 19–26. 26. Matz K, Keresztes K, Tatschl C, Nowotny M, Dachen- 34. Citrome LL, Holt RIG, Zachry WM, Clewell JD, Orth hausenm A, Brainin M and Tuomilehto J. (2006) Dis- PA, Karagianis JL and Hoffmann VP. (2007) Risk orders of glucose metabolism in acute stroke patients: of treatment-emergent diabetes mellitus in patients re- an under-recognized problem. Diabetes Care, 29 (4), ceiving antipsychotics. Annals of Pharmacotherapy, 41 792–7. (10), 1593–603. 27. Vancheri F, Curcio M, Burgio A, Salvaggio S, Grut- 35. Ramaswamy K, Masand PS and Nasrallah HA. (2006) tadauria G, Lunetta MC, Dovico R and Alletto M. Do certain atypical antipsychotics increase the risk of (2005) Impaired glucose metabolism in patients with diabetes? A critical review of 17 pharmacoepidemi- acute stroke and no previous diagnosis of diabetes mel- ologic studies. Annals of Clinical Psychiatry, 18 (3), litus. Q J Med, 98 (12), 871–8. 183–94. 28. Ostchega Y, Paulose-Ram R, Dillon CF, Gu Q and 36. Bellantuono C, Tentoni L and Donda P. (2004) An- Hughes JP. (2007) Prevalence of peripheral arterial dis- tipsychotic drugs and risk of type 2 diabetes: An ease and risk factors in persons aged 60 and older: data evidence-based approach. Human Psychopharmacol- from the National Health and Nutrition Examination ogy, 19 (8), 549–58. Survey 1999-2004. Journal of the American Geriatric 37. Shaw JE, Punjabi NM, Wilding JP, Alberti KG and Society, 55 (4), 583–9. Zimmet PZ. (2008) Sleep-disordered breathing and 29. Regensteiner JG, Hiatt WR, Coll JR, Criqui MH, type 2 diabetes: a report from the International Diabetes Treat-Jacobson D, McDermott MM and Hirsch AT. Federation Taskforce on Epidemiology and Prevention. (2008) The impact of peripheral arterial disease on Diabetes Res Clin Pract, 81 (1), 2–12. health-related quality of life in the Peripheral Arterial 38. Punjabi NM, Shahar E, Redline S, Gottlieb DJ, Disease Awareness, Risk, and Treatment: New Re- Givelber R and Resnick HE. (2004) Sleep-disordered sources for Survival (PARTNERS) Program. Vasc Med, breathing, glucose intolerance, and insulin resistance: 132 (1), 15–24. the Sleep Heart Health Study. American Journal of 30. Gorter PM, Olijhoek JK, van der Graaf Y, Algra Epidemiology, 160 (6), 521–30. A, Rabelink TJ, Visseren FL and the SMART 39. Willi C, Bodenmann P, Ghali WA, Faris PD and Study Group. (2004) Prevalence of the metabolic Cornuz J. (2007) Active smoking and the risk of type syndrome in patients with coronary heart disease, 2 diabetes: A systematic review and meta-analysis. cerebrovascular disease, peripheral arterial disease or Journal of the American Medical Association, 298 (22), abdominal aortic aneurysm. Atherosclerosis, 173 (2), 2654–64. 363–9. SECTION II Screening and Diagnosis

Simon Croxson United Bristol Healthcare NHS Trust, Bristol General Hospital, Bristol, UK

Here, we will consider the diagnosis of diabetes in Key messages the elderly – why a diagnosis should be made, the tests to be made for such diagnosis, and the types of diabetes • Undiagnosed diabetes is common in the older encountered. We will consider these points from the population. settings of older people at home in the community • Diagnosis is probably beneficial to the subjects. (‘free range’), of those in residential care, and of those • Elderly subjects with undiagnosed diabetes char- being admitted as hospital inpatients. acteristically have isolated post-challenge hyperglycaemia. • A non-diabetic fasting plasma glucose level does 3.2 Definition of diabetes not exclude diabetes, and glucose tolerance tests in the elderly are often required. Although many different diagnostic criteria for diabetes have previously been defined, they were rationalized 3.1 Introduction in 1979/1980 in reports from both the National Di- abetes Data Group (NDDG) and the World Health In this chapter, attention is focused on the diagnosis, Organization (WHO) [1, 2]. At that time, the ‘gold testing and screening of diabetes in the elderly, mainly standard’ definition was to wait for 3 to 10 years in relation to type 2 diabetes (T2DM). without any formal diagnosis or treatment; then, if Although type 1 diabetes (T1DM) does occur in diabetes-specific complications (generally retinopathy) the elderly, and is not uncommon, it is often clini- or overt diabetes symptoms were shown to be present, cally apparent that the problem with diabetes is not its a diagnosis of diabetes was made [3–7]. These stud- detection but rather the recognition of its type. Sec- ies showed that a 2 h post-glucose challenge (generally ondary diabetes is a disease of the elderly, often as a 75 g anhydrous glucose) venous plasma glucose level − result of chronic pancreatitis and steroid therapy rather of 11.1 mmol l 1 or more predicted the presence of than being secondary to other endocrine conditions or specific complications, and that ‘normal’ or elevated pregnancy; this presents in similar fashion to T2DM, fasting plasma glucose levels (using the criteria of the with problems being related to both the diagnosis and time) did not predict the absence of specific complica- recognition of type. tions [4, 6].

Diabetes in Old Age Third Edition Edited by Alan J. Sinclair © 2009 John Wiley & Sons Ltd. ISBN: 978-0-470-06562-4 22 CH 03 DIABETES IN THE ELDERLY: DIAGNOSIS, TESTING AND SCREENING However, as these studies were conducted predomi- between the OGTTs (as in Feskens’ study), the subjects nantly middle-aged populations, would the same crite- might adopt a more healthy lifestyle, or develop dia- ria apply to elderly subjects? betes. Feskens et al. used a 50 g glucose load, and cited If the prevalence of undiagnosed diabetes is ≥10%, evidence that such a load would induce minimal vari- then a histogram of the results of oral glucose tolerance ability. In contrast, others have shown higher glucose tests (OGTTs) would show a bimodal distribution of loads to be more consistent than smaller loads [19]. normal and diabetic groups, with a plasma level of Nonetheless, care must be taken to perform the OGTT 11.1 mmol l−1 dividing the two [4, 8–11]. This has correctly, including an overnight fast of 9 h or more, been shown not only in middle-aged populations, but no food or medication to be ingested on the morning also in elderly and old populations. It is believed, of the test, and only water to be drunk. In addition, no therefore, that the modern criteria apply in the elderly, exercise should be taken during the test, which should although no long-term follow- up studies of specific be performed in the morning (in general, a higher re- complications to confirm have been conducted, for sult is obtained during the afternoon). Several major, obvious reasons. incorrectly conducted surveys have resulted in preva- As will be noted later, the elderly diabetic person lences higher than reported in comparable surveys, as generally has an elevated post- challenge glucose level noted by Mykkanen and colleagues [20]. but a normal fasting level; this is termed ‘isolated post-challenge hyperglycaemia’. Whether this is a real 3.2.2 Other categories of glucose illness, and whether it matters, remains the subject intolerance of debate, with several population-based studies, as There are two other categories of glucose intolerance, well as the DECODE review, having shown isolated namely impaired glucose tolerance (IGT) and impaired post-challenge hyperglycaemia in elderly people to be fasting glucose (IFG): associated with an adverse outcome (often death) when compared to normal glucose tolerance [12–15]. • The concept of IGT was introduced in 1979/1980 [1, 2] as a statement of between what was re- 3.2.1 The oral glucose tolerance test garded as ‘normal’ and ‘diabetic’, with subjects hav- (OGTT) ing a higher risk of progressing to diabetes over time, in proportion to the prevalence of diabetes At present, the clinical gold standard for the diagnosis among the population. IGT is also associated with of diabetes is the OGTT, in which venous plasma large-vessel disease and an increased mortality rate glucose samples are measured. Although values for [21] (Table 3.1). Among the population of Asturas, other types of sample have been described, they seem aged 30–75 years, progression from IGT to dia- not to provide the same diagnoses [16]. betes was ascertained by conducting repeat OGTTs Previously, it has been reported by many that the at 6-year intervals; the independent risk factors for OGTT can be a ‘variable’ test [17]. For example, when progression were fasting and 2 h plasma glucose Feskens et al. repeated OGTTs in 237 subjects, aged levels, triglyceride levels and the body mass in- 64–87 years, annually between 1971 and 1975, the co- dex (BMI) [22]. Age, gender, education level, phys- efficients of intra-individual variation for fasting and ical activity, family history of diabetes, systolic 30, 60 and 120 min glucose ranged from 12% to 18% and diastolic blood pressures, total cholesterol and [18]. Such variability was not associated with age, HDL-cholesterol each also increased the risk of pro- gender, drug use or disease prevalence, although the regression, and were dependent on the preceding vari- liability coefficient was shown to depend on the preva- ables. lence of diabetes in the population, this being higher than was observed in younger populations. When clas- • Impaired Fasting Glucose (IFG) was first introduced sifying according to the WHO criteria, the variability in the 1997 American Diabetes Association (ADA) was comparable to that of other cardiovascular risk diagnostic criteria [23] and, as its name implies, is a factors, such as serum total cholesterol. state with a fasting plasma glucose between ‘normal’ The variability of the OGTT has been shown similar and diabetic. People with IFG should be subjected to that of other biochemical measures. A degree of the to an OGTT (despite the fasting level already having variability might in fact be due to variability among the been checked) to accurately clarify their glucose subjects; for example, should a prolonged time elapse tolerance status. 3.3 WHY DETECT DIABETES? 23 • Considerable interest has also been expressed in us- equivalent vascular morbidity and mortality to my- ing the fasting plasma glucose (FPG) or glycosylated ocardial infarction, and both have a 7-year vascular haemoglobin in lieu of the OGTT. Both, McCance death rate of 42% [36]. et al. and others have shown that the use of various • There is some evidence that a delayed diagnosis cut-offs for these values provides a good sensitivity leads to greater complications [38–40], but of course for predicting future retinopathy, though not quite as effective as the 2 h post-challenge value [24]. This this may be due to other factors such as lead time report was the subject of much constructive and crit- bias. The behaviour of those patients who delay ical correspondence [25–27], as it highlighted the seeking medical help might also lead them to harm. 5-year incidence of retinopathy (17.0% by raised 2 h • The identification of diabetes allows hyperglycaemia glucose; 11.3% by raised FPG; and 9.16% by raised to be controlled; this in turn would improve cogni- glycosylated haemoglobin). It was noted, however, tion and the quality of life [41]; it might also avoid that the high performance of these tests in Mc- microvascular complications [42]. Cance’s study was increased by performing the study in Pima Indians, a population with a naturally high • The identification of diabetes allows screening for prevalence of diabetes [28]. eye disease. Because a single raised plasma glucose level does • Identification should also reduce the risk of not serve as a diagnosis of diabetes (despite almost all metabolic decompensation; in the Rhode Island epidemiological surveys using just one test), confirma- study, hyperglycaemic coma was more likely to tory evidence is required in clinical practice; this might be fatal if the diabetes had not been previously be a further raised plasma glucose level at a differ- diagnosed [43]. ent time, and/or include osmotic symptoms or specific complications (e.g., retinopathy). • The identification of diabetes enables secondary care The latest WHO and International Diabetes Federa- providers to claim for more complex (expensive) in- tion (IDF) criteria [29] have simplified matters greatly patient episodes; early identification also allows an by introducing ‘intermediate hyperglycaemia’ as locat- easy attainment of Quality and Outcomes Frame- ing somewhere between a normal glucose tolerance work targets, increasing the UK GP’s income. (venous FPG <6.1 mmol l−1 and 2 h post challenge − • An accurate ascertainment of glucose tolerance sta- <7.8 mmol l 1) and diabetes (venous FPG ≥7.0 mmol − − tus is important not only for diabetes-specific re- l 1 and 2 h post challenge ≥11.1 mmol l 1). search, but also for other studies where diabetes is an important predictor variable. 3.3 Why detect diabetes? Although I believe that the detection of diabetes in older people is unquestionably valuable, no formal Although opinion varies as to whether diabetes should controlled trials have been conducted to confirm this be actively detected, or simply allowed to be presented proposal. At this point, it is worth considering the dif- in the clinical situation, detection may be preferred for ferent tests for diabetes in comparison to the OGTT a variety of reasons: using ‘modern’ criteria – that is, a 2 h post-challenge plasma glucose level of 11.1 mmol l−1 or more. Be- • Diabetes is a common condition, with its preva- fore the IGT concept was introduced in 1979/1980, lence varying depending on the age, ethnic origin, approximately 17% of all subjects in Rochester Min- residence and country of the population [30–35]. nesota who had been diagnosed diabetic in fact had Approximately 30–50% of diabetes cases are undiag- IGT rather than diabetes [44, 45]. One should examine nosed, such that between 3% and 15% of the elderly whether a simple test such as glycosuria or a raised population may have undiagnosed diabetes. FPG could be used to identify subjects in whom an • The identification of diabetes allows the aggressive OGTT is worth performing, rather than a screening management of vascular risk factors such as dys- test to instantly identify diabetic subjects (although this lipidaemia, hypertension and the use of antiplatelet may occur with some screening tests, such as a dra- agents, as diabetes represents Coronary Heart Dis- matically elevated FPG). In particularly, the settings ease equivalence [36, 37]. Having diabetes has an of the test – whether in the community, in residential 24 CH 03 DIABETES IN THE ELDERLY: DIAGNOSIS, TESTING AND SCREENING

Table 3.1 The 1999 WHO criteria for the diagnosis of diabetes. Values shown are glucose concentrations (mmol l−1). Diagnostic category Plasma glucose Whole-blood glucose Venous Capillary Venous Capillary Diabetes Fasting or ≥7.0 ≥7.0 ≥6.1 ≥6.1 2-h post load ≥11.1 ≥12.2 ≥10.0 ≥11.1 IGT Fasting and <7.0 <7.0 <6.1 <6.1 2-h post load 7.8–11.0 (incl.) 8.9–12.1 (incl.) 6.7–9.9 (incl.) 7.8–11.0 (incl.) IFG Fasting and 6.1–6.9 (incl.) 6.1–6.9 (incl.) 5.6–6.0 (incl.) 5.6–6.0 (incl.) 2-h post load <7.8 <8.9 <6.7 <7.8

Note: To convert mmol l−1 to mg dl−1, the value should be divided by 0.0555. care or in an acute medical admission – will be con- that they would seek – and receive – appropriate med- sidered ical attention. It is also important to be aware that, as the preva- In the Melton screening survey, among 24 elderly lence of an illness increases, then the performance of people aged 65–85 years found to have (previously un- any screening test will be improved. When studies are diagnosed) diabetes by OGTT, seven of them had thirst conducted in high-risk populations, such as Hispanics or polyuria on questioning. An absence of symptoms or Pima Indians of the USA, a screening test will per- was more likely in those subjects aged over 79 years form better than in a low-prevalence population [28]. and none of the eight octogenarians had symptoms A variety of other problems have been identified [49, 50]. Among 210 non-diabetic subjects, five had with many of the studies; for example, while some use thirst or polyuria for no obvious reason (non-diabetic a screening test and concentrate on those patients who OGTT, no diuretic use, normal renal function, cal- screen positive compared to just a handful of negative ‘screenees’, and elderly participants may be limited not cium and potassium levels). Interestingly, one of the only in numbers but also in age, often up to only 75 symptomatic subjects had consulted his GP and been years. reassured that there was not a problem; another subject denied osmotic symptoms, but had a glass of water in every room because he liked water. 3.4 The symptoms of diabetes In the Tampere screening survey [51], none of 19 elderly (aged 80+ years) new diabetic subjects was The classic osmotic diabetic symptoms of polyuria and found to have symptoms of diabetes. polydipsia occur when the renal threshold for glyco- Thus, in the ‘free range’ population, osmotic symp- suria has been exceeded. As the renal threshold of toms are specific but insensitive. glucose rises with age [46], it would be anticipated In the Birmingham Care Home Study [31], care that symptoms should occur quite late in disease pro- home residents received an OGTT and were asked gression. The questioning of newly presenting diabetic about their osmotic symptoms; among 44 found to have subjects of all ages at a diabetes centre revealed that diabetes, 22 had polyuria, 22 had thirst and 22 felt only 39% of patients had noticed symptoms of their diabetes, yet 80% had symptoms when directly ques- tired; the corresponding figures for 227 non-diabetic tioned [47]. When considering all age groups of the UK residents were 116, 116 and 121. Thus, among care general population, the elderly were considered more home residents, osmotic symptoms are insensitive and likely to recognize the symptoms of diabetes than were as non-specific as possible. the young [48]. Overall, it is hoped that the public will come to Very few screening surveys have investigated the recognize the significance of osmotic symptoms, and presence of symptoms, but of course if the subjects report to medical care and receive appropriate diagnos- were shown to be symptomatic then it might be hoped tic testing. 3.6 FASTING PLASMA GLUCOSE AND THE MODIFIED ORAL GLUCOSE TOLERANCE TEST 25 3.5 Glycosuria relatively specific (e.g., 90%). Urinalysis has the advantage that it provides a cheap and simple means In the Bedford diabetes survey, approximately half pf of selecting people for an OGTT, rather than simply the subjects with undiagnosed diabetes had postpran- testing the entire population with an OGTT; unfortu- dial glycosuria at all ages [52]; among undiagnosed nately, however, half of the subjects with undiagnosed diabetic subjects in the Islington diabetes survey, 33% diabetes will remain undiagnosed and be falsely reas- had fasting glycosuria and 73% had postprandial gly- sured. An example of this can be deduced from the cosuria [53]. approximate costs of the Ipswich and Melton stud- In the Melton diabetes survey in the elderly, gly- ies; if £900 (1988 prices) was available to spend, then cosuria was present in 13 of 25 diabetic subjects, in screening with OGTTs would identify one diabetic per- three of 20 IGT subjects, and in 12 of 225 normal son in the Melton survey, and miss none. In contrast, a subjects. Ignoring those subjects with known diabetes pre-screening with urinalysis would identify 10 diabet- and those who were not tested, and extrapolating to ics but would probably miss a similar number, despite those who had OGTT but no urinalysis, provides a screening a much larger population. sensitivity 53%, a specificity of 94%, and predictive values of positive and negative tests of 23% and 98%, 3.6 Fasting plasma glucose and the respectively [49, 50]. Very few other studies have been reported examining modified oral glucose tolerance glycosuria as a means of diabetes detection [54] using test post 1979/1980 criteria. Glycosuria was assessed in both Ipswich [55] and Ringkøbing, Denmark [56], Although, in the past the FPG has often been favoured where the population tested their urine and a small because of its simplicity, it has long been realized number of negative screenees were invited to take an that elderly subjects may have a clearly elevated 2 h OGTT. In both studies, the elderly tended to have post-challenge glucose level, despite non-diabetic FPG a higher participation rate than the young, and this levels [57, 58]. Based on population-wide information, difference was significant in elderly Danish men. this was apparent from the NHANES 2 data, which In the Ringkøbing study, when 106 subjects with- showed the average FPG to vary very little with age, out glycosuria were tested using a fasting capillary while the average post-challenge level tends to rise whole blood glucose, three were found to have dia- significantly with age [32]. betes [56]. Although, the prevalence of diabetes was There is a tendency to use the above data to support shown to be 9-11% for Danes aged 60–70 years by the premise that the FPG should be used for diag- other OGTT screening studies, the Ringkøbing screen- nostic purposes because, by using the post-challenge ing study showed only 0-2% of the elderly subjects value, too many ‘diabetic’ subjects would be identi- to have diabetes. Among all age groups, however, the fied. However, the original OGTT data indicated that sensitivity of urinalysis for glycosuria was calculated the incidence of specific complications was related to to be 20%. the post-challenge glucose levels, and that this occurred In the Ipswich study, OGTTs were conducted in peo- whether the FPG was raised, or not. This effect has re- ple with glycosuria and in 442 participants in the Isle of cently been confirmed in a series of larger, multi-ethnic Ely OGTT screening project; the two groups were sub- studies [59]. sequently amalgamated and the data analyzed together Data from Rancho Bernardo’s elderly population, [55]. This provided a sensitivity of 89% (albeit across NHANES 2 and the DECODE group has indeed shown subjects of all ages) and, by enriching the sample with that isolated post-challenge hyperglycaemia is associ- positive urinalysis subjects, made the test results appear ated with adverse outcome (death) compared to normal better than they in fact were. Friderichsen subsequently glucose-tolerant individuals of similar age from the calculated that a sensitivity of 17% would be more re- same populations [12–15, 60]. alistic for the Ipswich data [56]. Other studies from the DECODE group have shown Thus, very few studies have been conducted to as- that the FPG tends to be raised in fatter, younger sess the value of urinalysis in elderly populations com- patients, but not in older, slimmer patients [61]. pared to an OGTT; where performed, the sensitivity One further area of confusion is the difference be- was found to be low (e.g., 50%) but the test was tween a diabetic FPG (≥7.8 mmol l−1 from 1979, 26 CH 03 DIABETES IN THE ELDERLY: DIAGNOSIS, TESTING AND SCREENING

≥7.0 mmol l−1 from 1997) and a value above nor- and in testing people of different races aged 25 to 74 mal (6.0 mmol l−1 from 1997 onwards); one of these years in Newcastle, the prevalence of diabetes was is definitely a diabetic value (although confirmatory ev- higher in all three ethnic groups using the new fast- idence of the diagnosis is required clinically) and the ing ADA criteria compared to the WHO post-challenge other is not normal. criteria: 7.1% versus 4.8% in Europeans; 6.2% versus Subjects with a FPG that is above normal, but not 4.7% in Chinese; and 21.4% versus 20.1% in South in the diabetic range, are labelled as having IFG; this Asians [72]. Unfortunately, these are younger subjects is not regarded as an adequate diagnosis clinically, and are possibly overweight, thereby emphasizing the because in Bristol 30% of the subjects would have DECODE findings of different phenotypes with dif- diabetes on the 2 h values [62], while in the Leicester ferent abnormalities on OGTT in the young and old Gujerati population 85% would have diabetes [63]. [61]. Thus, subjects with IFG require their glucose level to Ultimately, this is all of historical interest, since be monitored post challenge (the fasting level for the the 1999 WHO and 2003 ADA criteria for diabetes OGTT has already been done). Further evidence that diagnosis are very similar, and the ADA now advocates subjects with non-diabetic FPGs can still not be normal the use of the OGTT [29]. comes from the Quebec Family Study, where subjects Recently, the ADA has altered its definition of ‘normal’; since the introduction of the 1979 National with non-diabetic OGTTs were divided into tertiles of Diabetes Data Group and the 1980 WHO criteria, the FPG [64]. Those in the highest FPG tertile had more USA and the rest of the world have used the same insulin resistance, less insulin secretion (corrected for criteria, apart from minor differences converting mg their insulin resistance) and more lipid abnormalities per deciliter to millimoles per litre. However, in 2003, than subjects in the lowest tertile. Also, within the the ADA decreased the normal FPG to under 5.6 mmol ‘normal range’ with post-challenge hyperglycaemia l−1 [73], while the WHO and International Diabetes excluded, a higher FPG is associated with an increased − Federation adhered to under 6.1 mmol l 1 [29]. mortality [65]. We recently examined the data from routine OGTTs In the Rancho Bernardo study of retirees aged 50–89 with regards to different FPG cut-offs in middle-aged years, only 69 of 254 newly diagnosed diabetic people and elderly subjects [74], the term ‘elderly’ was taken ≥ −1 had a raised FPG of 7.8 mmol l [58]. More as ≥60 years, following the WHO criteria [75]. Among usefully by modern criteria, Modan and colleagues 334 younger people tested, 121 had diabetes (FPG found that 49 of 134 newly diagnosed diabetic subjects 7.0+ mmol l−1 or post-challenge glucose 11.1+ mmol from a community survey of subjects aged 40–70 l−1); of these, one person (1%) had a FPG <5.6 mmol −1 years had a FPG below 6.4 mmol l [66]. In Pima l−1 and two people (2%) had a FPG of 5.6–6.0 mmol Indians of all ages, the screening of those not known l−1. Among 265 older people tested, 151 had diabetes; to have diabetes by OGTT showed that 8% of the of these, 10 (7%) had a FPG <5.6 mmol l−1 and 11 subjects with previously undiagnosed diabetes had a (7%) had a FPG of 5.6–6.0 mmol l−1. In young peo- −1 FPG under 6.1 mmol l [67], with corresponding ple, changing the normal plasma glucose value has figures for Taiwanese of all ages of 27% [68] and for little effect, as only 2.5% (95% CI: 0.5–7.0%) with Australians of all ages of approximately 30% [69]. undiagnosed diabetes have a FPG <6.1 mmol l−1.On In elderly Koreans with previously undiagnosed di- the other hand, again older people with undiagnosed abetes, 50% of the men and 40% of the women had diabetes often had a low FPG, with 10 (7%; 95% − a FPG under 6.1 mmol l 1 [70] and in elderly African CI: 3–12%) having a value <5.6 mmol l−1 and 21 − Americans 30% had a FPG under 7.0 mmol l 1 [71]. (14%; 95% CI: 9–21%) a value <6.1 mmol l−1;that Thus, the FPG in free range individuals and elderly is, 7% (95% CI: 4–13%) of undiagnosed elderly di- misses many people with undiagnosed diabetes. abetic subjects had a FPG of 5.6 to 6.0 mmol l−1, The over-riding difference between the ADA crite- inclusive. Thus, in young people, the FPG is reason- ria and the WHO criteria has been the ADA’s pre- able at excluding diabetes, whichever ‘normal’ FPG is occupation with using the FPG. Not surprisingly, the used. However, in older people the lower ADA cut-off prevalence of diabetes in the elderly Korean is approx- of <5.6 mmol l−1 as ‘normal’ is needed to detect many imately twice that by WHO criteria than by ADA cri- of the diabetic subjects (although even this will still teria [70], and similar findings occur in elderly African miss some subjects with undiagnosed diabetes). Un- Americans [71]. However, this is not always the case, fortunately, among the 151 diabetic elderly subjects, 3.7 RANDOM AND POSTPRANDIAL PLASMA GLUCOSE 27 29 (6%: 95% CI: 13–25%) did not have a diabetic 97%. Thus, many care home residents with undiag- 2 h post-challenge value but rather had a diabetic FPG nosed diabetes will have a low FPG; for example, − level. Consequently, it would seem that both fasting 40% had a FPG <5.0 mmol l 1 and only 18% had − and post-challenge glucose values are required for a a FPG ≥7.0 mmol l 1. The converse is that the mod- comprehensive assessment of glucose tolerance status ified oral glucose tolerance test (MOGTT) using just in older people. the 2 h post-challenge value is extremely useful; among In the Birmingham Care Home Study, the fasting the Birmingham Residential Home subjects, 43 of the capillary plasma glucose data were acquired alongside 46 new diabetic subjects had post-challenge hypergly- the OGTT results [31, 76]. From this, it is possible to caemia, often with unremarkable fasting glucose levels. draw a receiver operator characteristics (ROC) curve This has a sensitivity of 0.93, a specificity of 1.0, pre- (this may be drawn using statistical software) in which dictive value positive test of 1.0 and predictive value sensitivity is plotted against 1-specificity for all possi- negative test of 0.99. ble cut-offs (Figure 3.1). It is important to emphasize that the aim here is to identify the optimum FPG as 3.7 Random and postprandial a diabetes screen, and not to redefine the criterion plasma glucose for a diabetic FPG. Ideally, the ROC curve for the data would include a point with maximum sensitivity Postprandial glucose levels were comprehensively as- and specificity (i.e., at the top left-hand corner of the sessed by Engelgau et al. in Egypt [78], when a total of graph), which would also have the largest area under 828 subjects had a random capillary whole blood es- the ROC curve and may be a useful value to compare timation timed relative to their last meal, followed by screening tests (again the statistics software performs an OGTT. This was a very practical ‘real world’ study the calculations). A clear example is in the Coventry using a One Touch II meter, testing whenever possible Study, where the optimum cut-off for a random blood (but often afternoon) and with a variable time after a glucose to find diabetes was examined [77]. variable meal. By using an ROC curve, the random In the Birmingham Care Home Study, the optimum blood glucose level was found to have an optimum − fasting capillary plasma glucose was ≥5.8 mmol l 1, sensitivity with shorter postprandial periods, while the with sensitivity 42% and specificity 96%; the corre- cut-off level needed to decrease with longer postpran- sponding values for cut-offs for 5.0 mmol l−1 were dial periods yet to increase with advancing age. Thus, 61% and 70%, and for 6.0 mmol l−1 were 35% and for subjects aged 75 years the optimal cut-off was

Sensitivity 1.00

0.75

0.50

0.25

0.00 0.00 0.25 0.50 0.75 1.00 1-Specificity

Figure 3.1 Receiver operator characteristics (ROC) curve from the Birmingham Care Home Study of fasting plasma glucose for diabetes diagnosis 28 CH 03 DIABETES IN THE ELDERLY: DIAGNOSIS, TESTING AND SCREENING

7.8 mmol l−1 at 1 h postprandial, with a sensitivity of unknown how many would have surfaced naturally if 81% and specificity of 80%. Again, the study results left to their own devices. emphasized that different cut-offs were required in the The use of a random blood glucose measurement young and old; for example, for 30-year-old subjects was investigated in the Birmingham Care Homes Study the optimal level was 1 h postprandial, with a glucose using BM-test 1–44 test strips and Reflolux S meter level of 6.55 mmol l−1, and a sensitivity of 82% and [Boehringer Mannheim (BM), Mannheim, Germany]. specificity of 78%. An ROC curve analysis gave an optimum BM test −1 In the Coventry Study, the bi-ethnic population of cut-off of 5.3 mmol l [31, 76], with a sensitivity of Coventry was prescreened with a random glucose level, 56% and a specificity of 82%. Given the findings in ≥ −1 and then subjected to OGTT if screen-positive (capil- Egypt above and Bristol below, an 8.0 mmol l lary whole blood glucose ≥6.0 mmol l−1 within 2 h cut-off was examined that showed a sensitivity of 4% of meal, or ≥5.0 mmol l−1 at more than 2 h postpran- and a specificity of 100% (it is standard finding that dially); a group of screen-negative subjects was also either sensitivity or specificity is gained at the expense subjected to an OGTT [34, 77, 79]. Interestingly, there of the other). was a significant difference in sensitivity for a ran- Previously, the use of a random venous plasma glu- dom capillary whole blood glucose cut-off of 8.0 mmol cose for diabetes detection in elderly medical admis- l−1 between elderly Europeans (sensitivity 36%) and sions to the hospital in Bristol, has been investigated. elderly South Asian subjects (sensitivity 68%) (the sen- In this predominantly white, European population [82], sitivity values for a 7.0 mmol l−1 cut-off were 50% and the ultimate prevalence of diabetes was 1% if the admission plasma glucose (APG) was 7.0–7.9 mmol l−1, 75%, respectively), while the addition of testing within 10% if it was 8.0–12.9 mmol l−1, and 50% if ≥13.0 2 h of a meal improved the sensitivity in South Asian, mmol l−1. Other interesting findings were that, in pa- but not European, people. At all cut-offs, the sensitiv- tients found to have diabetes as an inpatient, 50% were ity was greater in South Asian subjects than Europeans, non-diabetic on retesting at 6 weeks after discharge. presumably due to the much higher prevalence. How- The lowest APG found to be positive for diabetes was ever, very few elderly screen-negative subjects [43] 7.0 mmol l−1, while the highest APG to be negative were available, and the two subjects who had undi- for diabetes was >20 mmol l−1. As a consequence, agnosed diabetes but were screen-negative were both any patient admitted to the Bristol hospital with an elderly. − APG ≥8.0 mmol l 1 will be screened; in this way Again, there are other community-based studies us- one subject with undiagnosed diabetes will be missed ing random plasma glucose (RPG) estimations, and every per 6 months. Unfortunately, these tests are Andersson’s study is such an example [80]. Here, sub- not 100% accurate, and a diagnosis of diabetes made jects aged up to age 79 years and with a positive during a hospital admission should be reviewed at a urinalysis for glucose or random capillary whole blood later date. ≥ −1 glucose 8.0 mmol l were followed up, with a fast- Thus, while a RPG in a care home resident is ing whole blood glucose (FBG) assessment; if the FBG generally unhelpful, within the free range population was neither diabetic nor very low (FBG 5.5–6.6 mmol a 1 h postprandial plasma glucose ≥7.8 mmol l−1 or, −1 l , inclusive), an OGTT was performed; as a conse- in the acute hospital admission a RPG ≥8.0 mmol quence, 3268 people (85% of the target population) l−1, represent useful tests to determine which patients were tested over 3 years, and 66 were found to be di- should receive an OGTT to check for diabetes. abetic. Urinalysis detected 20 diabetic subjects, while blood glucose testing detected 64; other similar studies confirmed the much higher sensitivity of blood testing 3.8 Blood glucose meters over urine testing [81]. The drawbacks to this study were that it excluded subjects aged >79 years, there Modern studies measuring venous plasma glucose use was no assessment of screen-negative subjects, and it a glucose oxidase method which is specific for glu- was not possible to compare the final prevalence in cose. However, meters for self glucose monitoring the general practice population to that of a screened measure whole blood glucose, and either report whole population. On the other hand, the results showed that blood glucose or report a plasma glucose equivalent a keen general practice can use a screening system to [83]. Occasionally, in fact, it is difficult to deter- identify many (although not all) diabetic subjects. It is mine from the manufacturer exactly what is being 3.9 GLYCOSYLATED HAEMOGLOBIN 29 reported by the meter. Typically, whole blood glu- study, 232 subjects with FPG <6.4 mmol l−1 showed −1 cose is approximately 10–15% (1 mmol l ) lower than an increase in HbA1c with age [90], although it was plasma glucose due to the presence of red blood cells; not known how many of these subjects would have hence, anaemic subjects will record higher readings isolated post-challenge hyperglycaemia, which is more on whole blood glucose measurements, as are used in common with age. Others have shown that, in a group self blood glucose meters. The plasma/whole blood dif- of 93 subjects, age was not associated with any change ference depends on the haematocrit; a 15% decrease in FPG or GHb [91], although this size sample would in glucose level is evident when the haematocrit is have been highly prone to a type 2 error. Thus, there is 0.55, but only 8% when the haematocrit is 0.31 [83]. still scope to determine whether older folk with normal Thus, attention must be paid to the meters used in the glucose tolerance have a significantly higher HbA studies, as well as remaining aware of other factors 1c than do younger folk. that might influence the results obtained with glucose GHb has been assessed as a tool for diabetes diag- meters. nosis in many studies, and is indeed an almost standard It is also important to remember that fingerprick capillary glucose values may be falsely low in sick investigation in such trials. patients, and this may lead to an underestimation of An analysis of data from a large database of their hyperglycaemia [84, 85]. the Meta-analysis Research Group (MRG) and the NHANES III data (subjects aged 40–74 years) showed that two-thirds of subjects with a 2 h post-challenge 3.9 Glycosylated haemoglobin plasma glucose of 11.1–13.3 mmol l−1 had a normal HbA1c (up to 6.3%), while only one-third with an −1 Glycosylated haemoglobin (GHb) assays have been OGTT result >13.3 mmol l had a raised HbA1c [92]. developed and enhanced over the past three decades, The data from the MRG study over-represented sub- and now measure the more specific HbA1c than the jects with undiagnosed diabetes, which again makes original HbA1; there are, however, several different screening tests perform better, although the sensitiv- assay methods which can produce slightly different ities were low. The finding that an elevated HbA1c results. In addition, a variety of factors can influence missed many diabetic subjects and only identified those GHb other than plasma glucose [86]. For example, a with an elevated HbA1c, was described as being advan- rapid turnover of red blood cells due to haemolysis, tageous, as it only finds those diabetic subjects with sickle cell disease or thalassemia will reduce GHb raised HbA1c in whom an intervention would be made levels, whereas the persistence of fetal Hb will cause to lower such levels (a rather self-recursive argument); an increase in GHb. In particular, the effect of variant this clearly misses the majority of diabetic subjects haemoglobins differs with different assays. in whom aggressive intervention would be made for Interestingly, some subjects are higher ‘glycators’ vascular protection. than others; this is unrelated to glycaemic levels, and A recent systematic review of HbA compared to most likely due to a longer red cell survival in high 1c formal OGTT was prepared by Bennett et al. [93]. glycators. It may lead to a GHb difference of 2% Here, the most interesting point was that only nine between high and low glycators with normal glucose studies met the inclusion criteria which were that the tolerance [87]. It is often written that the GHb normal range in- study had to be written in English, the majority of creases with age. In the France Telecom project, work- subjects had to have had OGTT by modern criteria, and the HbA1c result could be aligned to Diabetes Control ers were screened using OGTT and HbA1c measurements. Among 3240 subjects aged 18 to 80 years, and Complications Trial values. The results showed very few had abnormal glucose tolerance (22 diabetic, HbA1c to have a slightly lower sensitivity than FPG 210 IGT subjects) and the HbA1c was seen to rise in detecting diabetes, but a slightly higher specificity. ≥ significantly with age, before falling in males [88]. At an HbA1c cut-off of 6.1%, the sensitivity ranged Certainly, splitting the OGTT data into standard cate- from78to81%,andspecificityfrom79to84%.For − gories, rather than treating it as a continuous variable FPG at a cut-off point of ≥6.1 mmol l 1, the sensitivity [89], would have lost some of the power of the data ranged from 48 to 64% and specificity from 94 to 98%. to detect whether age itself or an increase in glucose Both, HbA1c and FPG have a low sensitivity for the levels with age cause this rise in HbA1c. In a further detection of IGT (ca. 50%). 30 CH 03 DIABETES IN THE ELDERLY: DIAGNOSIS, TESTING AND SCREENING

In the Melton diabetes survey the HbA1 levels were The attraction of fructosamine is that its assay is easily measured, using a Corning Gel electrophoresis method, automated, which makes it less costly. However, fruc- in subjects with diabetes and in a random selection of tosamine levels are affected not only by the glucose other subjects with normal and impaired glucose toler- level and the assay method, but also by albumin and ance, with an upper limit normal range of 8.5% [50]. lipid levels, and the time of day when the sample is Fifteen of 25 new diabetic subjects had an elevated taken [95]. HbA1, while only one of 26 normal glucose-tolerant In the Melton study, fructosamine levels were mea- subjects had a raised HbA1. The extrapolation of these sured in 264 normal glucose-tolerant subjects using a data to the whole population, including IGT subjects, standard method with the Cobas biocentrifugal ana- provided a sensitivity of 63% but a specificity of 91% lyzer, standardized against a human albumin solution In the Birmingham Care Home Study, HbA1c was [50, 96]. The upper limit of the normal range was taken measured using a latex-enhanced competitive turbidi- as the 95th centile (1.92 mmol l−1), which gave a sen- metric immunoassay (Unimate 5 HbA1c; Roche Diag- sitivity of 74%, and specificity and specificity of 95%. nostics Division, F. Hoffmann-La Roche Ltd, Basel, However, the authors noted that the upper limit of the Switzerland) with a reference range of 3.5–6.5% [31, ‘normal’ range varied widely, from 1.18 to 3.12 mmol −1 76]. When considering an HbA1c elevated above the l in different populations reported elsewhere. normal range (>6.5%), 10 of 38 diabetic subjects and In recent years the fructosamine assay has been 12 of 189 non-diabetic subjects had elevated HbA1c improved by removing other substances that interact levels (sensitivity 26%, specificity 94%, predictive with the assay, and by using improved standardization value positive test = 0.45, predictive value negative [97]. test = 0.86). The ROC curve analysis gave an optimum Among 157 free range elderly subjects (aged cut-off of 6.2%, with elevated HbA1c in 19 diabetic 65–88 years), serum fructosamine levels (first- and and 32 non-diabetic residents (sensitivity 0.5, speci- second-generation assays) were monitored and OGTTs ficity 0.83, predictive value of positive test = 0.37, performed [98]. As a consequence, 16 undiagnosed predictive values of negative test = 0.89). asymptomatic diabetic subjects were identified. The use of HbA1c (DCCT aligned) has been studied ROC curves were used to provide the optimum − in acutely admitted stroke patients; this included all cut-offs. When using a cut-off ≥2.3 mmol l 1 for subjects who had an admission plasma glucose level the first-generation assay, the sensitivity to detect ≥6.1 mmol l−1 who were later subjected to an OGTT diabetes was 75%, specificity 83%, and positive μ at 12 weeks [94]. Here, a HbA1c level of ≥6.2% had a predictive value 35%. With a cut-off of 250 mol − sensitivity of 86% and a specificity of 94%. It would be l 1 for the second-generation assay, the sensitivity expected that few subjects with undiagnosed diabetes to detect diabetes was 81%, specificity 87%, and would have an admission plasma glucose <6.1 mmol positive predictive value 43%. Consequently, while −1 l . Hence, in this study the HbA1c test performed the two fructosamine assays performed similarly in well, possibly due to the high underlying prevalence the elderly, the enhancement of the assay contributed of diabetes (21%) in this group. little, as did adjustments for serum albumin or protein Although, in older people HbA1c is poorly sensitive, levels in other studies. it is specific for undiagnosed diabetes. Yet, this may It may be useful to combine several blood tests. prove beneficial in the clinical situation, when an ill For instance, in a Hong Kong population, the paired −1 patient has hyperglycaemia: if their HbA1c level is values of an FPG of 5.6 mmol l and a HbA1c of raised, they are highly likely to have diabetes; however, 5.5% gave a sensitivity of 83.8% and specificity of if it is not raised they might still be diabetic. 83.6% to predict a 2 h plasma glucose of ≥11.1 mmol l−1, while the paired values of an FPG of 5.4 mmol l−1 and a fructosamine level of 235 μmol l−1 gave an 3.10 Fructosamine optimal sensitivity of 81.5% and specificity of 83.2% [99]. Although these figures were impressive, again the Fructosamine is misleadingly named as it consists of population was selected to be at high risk of glucose glycosylated plasma proteins (mainly albumin), and intolerance, which improves the performance of the does not involve fructose at all. Current clinical ex- diagnostic tests; nonetheless, the result proved to be perience with fructosamine is far inferior to that with excellent in these subjects. However, these findings GHb, and the measurement is less well standardized. should be counterbalanced by Modan’s study in an 3.12 KNOWN DIABETES 31 Israeli general population, the results of which showed lower than originally described, which was mainly that the addition of HbA1 to the FPG contributed due to population variation of risk factors compared zero [57]. to the KORA sample (age, BMI, antihypertensive medication and smoking). The Atherosclerosis Risk in Communities cohort 3.11 Diabetes prediction calculators study examined 7915 participants aged 45–64 years who were free of diabetes at baseline, and ascertained In recent years there has been a growing interest in 1292 incident cases of diabetes by clinical diagno- using patient information to select subjects at high sis or OGTT [105]. A risk function based on waist, risk of diabetes and in their screening, rather than to height, hypertension, blood pressure, family history of pre-screen with a blood or urine test. Examples include diabetes, ethnicity and age performed similarly to the the Finnrisk calculator [100] and the USA Diabetes fasting glucose level (AUC 0.71 and 0.74, respectively; Prevention Program calculator [101], both of which P = 0.2). The addition of fasting glucose to the clini- were developed for populations with a high risk of cal model improved the AUC marginally (to 0.78), and diabetes. It has been shown, however, that calculators including triglycerides and HDL-cholesterol provided that function well in one population often do not do so an even better performance (AUC 0.80; P < 0.001). in a different population. These models achieved sensitivities of 40–87% and The San Antonio clinical model formula used age, specificities of 50–86% – similar to when using a di- gender, ethnicity, fasting glucose level, systolic blood agnosis of metabolic syndrome. pressure, HDL-cholesterol level, BMI and parental or In the NHANES 3 study, which was conducted sibling history of diabetes [102]. In this predominantly intheUSAinsubjectsagedupto75years,a middle-aged Hispanic population the model predicted Diabetes Risk Calculator that included questions the development of diabetes over the next 7.5 years, on age, waist circumference, gestational diabetes, and probably worked well because the population was height, race/ethnicity, hypertension, family history and at very high risk of developing diabetes. When the San exercise, was developed [106]. This had sensitivity, Antonio formula was applied to Japanese Americans specificity, positive and negative predictive values, of all ages, the model was good at predicting diabetes and ROC for detecting undiagnosed diabetes of 88%, development in subjects under the age of 55 over a 75%, 14%, 99.3% and 0.85, respectively. 5-year period, but was not helpful in subjects aged The Rancho Bernardo population was used to derive over 55 years [103]. a prediction tool for abnormal glucose tolerance based Four screening tests – the Rotterdam Diabetes on age, gender, FPG and triglyceride levels, where a Study, the Cambridge Risk Score, the San An- cut-off of 4 points on the equation had a sensitivity of tonio Heart Study and the Finnish Diabetes Risk 46% and a specificity of 83% [107]. This was derived Score – were performed in 1353 participants (aged in a retired white population, but similar values were 55–74 years) without known diabetes in the Coop- found in white and black populations aged 70–79 years. erative Health Research in the Region of Augsburg Thus, while many prediction formulae may become (KORA) Survey, and compared to OGTT results available for the prediction of diabetes or abnormal [104]. The sensitivity, specificity and the area under glucose tolerance, they are highly dependent on the in- the ROC curve (termed the AUC) for undiagnosed dividual population being studied. Their performances diabetes were calculated. The AUCs were 61% (95% are clearly altered by the age and ethnicity of the sub- CI: 56–66%) for the Rotterdam Diabetes Study, jects such that, in some cases, it is no better than FPG. 65% (95% CI: 60–69%) for the Finnish Diabetes In fact, some other models actually include the FPG Risk Score (P = 0.10 versus Rotterdam), and 67% within the formula. (95% CI: 62–72%) for the Cambridge Risk Score (P < 0.001 versus Rotterdam). The San Antonio Heart Study model, which includes the fasting glucose level, 3.12 Known diabetes yielded an AUC of 90% (P < 0.01 versus all three questionnaires); however, this was not significantly It must be appreciated that the diagnosis of diabetes is different from fasting glucose level alone (AUC, not always correct; in the Melton diabetes survey, for 89%; P = 0.46). The sensitivities, specificities and example, 48 subjects were diagnosed as diabetic and predictive values of questionnaires were substantially had confirmatory blood tests in their medical records. 32 CH 03 DIABETES IN THE ELDERLY: DIAGNOSIS, TESTING AND SCREENING However, a further three subjects had a diagnosis T2DM, may rapidly fail if prescribed oral agents, and of diabetes which was incorrect on diabetes testing; the need for insulin must not be overlooked. among these three, one subject had impaired glucose As well as T1DM having an insidious onset in tolerance, one had a low renal threshold for glycosuria, the elderly, T2DM can – at all ages – present with and one simply had a name similar to another person ketosis resembling T1DM, particularly in non-white who did have diabetes [30, 50]. ethnic groups; this is known by many names, of In the Bristol admission plasma glucose study [82], which ‘ketosis-prone T2DM’ seems most appropriate. 70 subjects were ‘known’ to have diabetes, but on Umpierrez and colleagues recently published an ex- testing this was refuted in five cases. tremely succinct, yet comprehensive, review of the In the NHANES II study, a selected subgroup of subject [113]. Features suggestive of ketosis-prone ‘known diabetic’ subjects was tested, and 19% were T2DM include non-white ethnic origin, newly di- shown not to have diabetes [32]. agnosed diabetes, obesity, family history of T2DM, As well as the causes given above, other causes negative autoantibodies (e.g., GAD or islet cells), include stress-induced hyperglycaemia, drug-induced fasting C-peptide levels of ≥0.33 nmol l−1 within hyperglycaemia and incorrect application of the diag- one week after resolution of diabetic ketoacidosis nostic criteria. (DKA), or ≥0.5 nmol l−1 after 6–8 weeks, and a So, in a subject with ‘known diabetes’, the diagnosis glucagon-stimulated C-peptide level of 0.5 nmol l−1 at should be questioned, although continuing with the presentation and 0.75 nmol l−1 during follow-up. incorrect diagnosis does help to achieve glycaemic Some subjects with T2DM are relatively slim, and targets. are diagnosed with T2DM by virtue of a lack of autoantibodies; however, on testing their main defect is β-cell dysfunction and not insulin resistance [114]. 3.13 Types of diabetes T2DM with predominantly impaired β-cell function is particularly common among Oriental people. It is clear that the majority of diabetes in the elderly is Secondary diabetes is a disease of the elderly in T2DM. For example, in the Oxford study 95% of sub- which the incidence increases dramatically with age jects known to have diabetes had T2DM, as did 75% of [44]; although this may be due to altered endocrine the insulin-treated patients [108]. Nonetheless, T1DM conditions, in practice it is due to chronic pancreatitis, does occur in the elderly, with studies conducted in and also to diabetogenic drugs such as glucocorticos- Denmark and Rochester, USA, having shown that the teroids, thiazides, beta-receptor agonists (e.g., salbuta- incidence is similar at ages 30 to 80 years [44, 109]. mol tablets), β-blockers and the atypical antipsychotics The standard clinical criteria for classification as T1DM are either spontaneous significant ketosis, or [1, 115]. any two of personal history of autoimmune disease, Although, it may be difficult to decide at the time T1DM in a first-degree relative, significant osmotic of presentation of DKA, which type of diabetes a pa- symptoms, or significant weight loss (even if over- tient has, the recent Aβ classification [116] can be weight at onset and still overweight) [110]. extremely helpful, by using immunologic and β-cell In some subjects, autoimmune diabetes can have an function measures (Table 3.2). This examines the au- insidious onset, and is known as ‘latent autoimmune toantibodies and whether the β-cell function is pre- ≥ diabetes of the adult’ (LADA); the criteria include age served by virtue of a fasting serum C-peptide level 1 −1 ≥35 years, a 6–12 months period of diabetes prior ng ml or maximum glucagon-stimulated C-peptide −1 to insulin initiation and the presence of glutamic acid ≥1.5ngml , measured both at the time of DKA and decarboxylase (GAD) antibodies [111]. These subjects at 6–12 months later; usefully, the 6–12 month results have an HLA genotype between T1DM and T2DM, are often equivalent to the acute result, while the fast- their β-cell function deteriorates faster than in T2DM ing result is often equivalent to the stimulated value. but slower than in T1DM, and they tend not to have The significance of GAD positivity is still uncertain, the features of the metabolic syndrome. Although some however, since in a small group of elderly diabetic question the rationale for the designation of LADA subjects GAD-positive individuals showed a greater [112], it is important to realize that an elderly person insulin response than did GAD-negative individuals presenting with diabetes that has the appearance of [117]. 3.16 CONCLUSIONS 33

Table 3.2 Diagnostic outcomes of Aβ classiﬁcation in USA subjects with DKA [116]. β-Cell reserve present β-Cell reserve absent Antibodies −ve Ketosis-prone T2DM: 55% of population Insulin-requiring DM: 21% of population Antibodies +ve 1/2 T1DM 1/2 T2DM: 5% of population T1DM: 19% of population

Adapted from Balasubramanyam et al. Diabetes Care. 2006; 29(12): 2575–9.

3.14 Metabolic syndrome 3.15 Future research

This condition is the clustering of obesity, insulin Although, for future investigations, the use of resistance, dyslipidaemia and hypertension; previously, HbA1c as a screening tool might not be worthwhile, this has also been referred to as Reaven’s syndrome, it would be intriguing to determine whether this syndrome X and the ‘deadly quartet’. Its significance is parameter is affected by patient age. Likewise, while that it predicts a doubling of the risk of vascular events second-generation fructosamine assays may be of and also multiplies the risk of developing diabetes in interest as a screening tool, in large studies they non-diabetic subjects fivefold [118]. Some studies have would most likely perform in similar fashion to the suggested that the increasing risk of vascular events HbA1c assay. with increasing number of components of metabolic It might also be important to ascertain whether syndrome is additive, whilst others have suggested that diabetes detection in frail care home residents would the effect is greater [118]. lead to interventions that were beneficial, and this could The criteria for the diagnosis of metabolic syn- perhaps be achieved via an audit. drome from bodies such as the WHO and National Patient characteristics might also be used, either to Cholesterol Education Program – Third Adult Treat- provide a score suggesting that an OGTT should be ment Panel, have been evolving and converging over performed in the short term, or which other test(s) to recent years, and have culminated in the recent Interna- perform. For example, whereas very thin people re- tional Diabetes Federation (IDF) consensus guidelines quire the post-challenge glucose level to be measured, [118]. The key component of metabolic syndrome is cen- the FPG may suffice in an obese patient. Other impor- tral obesity, assessed by waist measurement (with IDF tant variables to consider in this model would be the cut-offs of ≥94 cm for Europid men and ≥80 cm for type of patient (e.g., care home resident, free range or Europid women); the IDF cite waist cut-offs for differ- acute hospital admission) and the BMI. ent ethnic groups and gender, which should be applied It is clear that paying more attention to postpran- to subjects wherever they live. The IDF committee dial glucose levels would be valuable, as might the make the point that the waist cut-offs are pragmatic, combination of variables such as FPG and HbA1c or from various sources, and should be refined with time; fructosamine. there is no consideration of change in waist measure- Finally, further studies of the Aβ classification in ment with age. other population groups to ensure the most appropriate The IDF definition of the metabolic syndrome [118] treatment for subjects presenting with DKA would not consists of central obesity (a waist circumference only be of interest but also of great therapeutic value. greater than ethnic-specific cut-offs, or BMI >30), plus any two of: raised triglycerides (>1.7 mmol l−1);low HDL-cholesterol (<1.03 mmol l−1 in males, 1.29 mmol 3.16 Conclusions l−1 in females); raised blood pressure (>129 mmHg systolic or 84 mmHg diastolic, or treated hypertension); The detection of diabetes in elderly people is clearly or a raised fasting plasma glucose (>5.5 mmol l−1 or worthwhile, not only because the undiagnosed disease diabetes diagnosed). is common but also that treatment can benefit these The management of metabolic syndrome is via patients in several ways. However, no controlled trials lifestyle modulation and the administration of drugs to have been conducted to justify this view. avoid the development of diabetes and adverse vascular A full ascertainment of glucose tolerance status re- events quires a two-point OGTT to be performed, but in 34 CH 03 DIABETES IN THE ELDERLY: DIAGNOSIS, TESTING AND SCREENING routine clinical practice it is impractical to test ev- 2. WHO Study Group. (1980) Technical Report Series ery potential patient. Thus an initial test is required 646: Diabetes Mellitus. World Health Organization, to screen for those subjects at high risk of diabetes Geneva. and, if identified, subsequently to perform the OGTT. 3. Sayegh H and Jarrett R. Oral glucose-tolerance tests While some subjects will be found to have glycosuria and the diagnosis of diabetes: results of a prospective on urinalysis, a few will show osmotic symptoms. Al- study based on the Whitehall survey. Lancet 1979, 2 though these features are less common with advancing (8140), 431–3. age, public education and opportunistic urine testing 4. Dorf A, Ballintine E, Bennett P and Miller M. might allow the identification of those people likely to Retinopathy in Pima Indians. Relationships to glu- have diabetes and in whom further testing is required. cose level, duration of diabetes, age at diagnosis of However, the evidence acquired from residential home diabetes, and age at examination in a population with settings suggests that osmotic symptoms are meaning- a high prevalence of diabetes mellitus. Diabetes 1976, less, although it would still be necessary to conduct 25(7), 554–60. a blood glucose test in the presence of any relevant 5. Jarrett R and Keen H. Hyperglycaemia and diabetes symptoms (i.e., in half of all residents). At present, ran- mellitus. Lancet 1976, 2 (7993), 1009–12. dom blood/plasma glucose measurements (notably 1 h 6. Saad M, Knowler W, Pettitt D, Nelson R, Mott D and postprandial) offer the best sensitivity and specificity Bennett P. The natural history of impaired glucose in both free range and acute hospital admission set- tolerance in the Pima Indians. N Engl J Med 1988, tings, but perform poorly in the residential care setting, 319 (23), 1500–6. where the modified OGTT shows superior sensitivity 7. Pettitt D, Knowler W, Lisse J and Bennett P. De- and specificity but it is complicated to carry out. velopment of retinopathy and proteinuria in relation It is important to appreciate that the above details ap- to plasma-glucose concentrations in Pima Indians. Lancet 1980, 2 (8203), 1050–2. ply to the different populations studied; with different 8. Rushforth N, Bennett P, Steinberg A, Burch T and ages, diabetes prevalences and/or BMIs, these screen- Miller M. Diabetes in the Pima Indians. Evidence of ing tests are likely to perform differently. The find- bimodality in glucose tolerance distributions. Diabetes ings of any pre-screening test will vary depending on 1971, 20 (11), 756–65. the population characteristics (notably age, residence, 9. Zimmet P and Whitehouse S. The effect of age on BMI), the technical details of the test used, and the glucose tolerance. Studies in a Micronesian population prevalence of diabetes in the population, with screening with a high prevalence of diabetes. Diabetes 1979, 28 tests performing better in high-prevalence populations. (7), 617–23. When considering a screening program, it is 10. Lim T, Bakri R, Morad Z and Hamid M. Bimodality also necessary to consider not only the scientific in blood glucose distribution: is it universal? Diabetes aspects – whether a specific population is being Care 2002, 25 (12), 2212–17. tested – but also the practical aspects. For example, 11. Fan J, May S, Zhou Y and Barrett-Connor E. although a simple screening test might only show 50% Bimodality of 2-h plasma glucose distributions in sensitivity, it would be much more easy to perform in whites: the Rancho Bernardo study. Diabetes Care all patients than would a two-point OGTT. 2005, 28 (6), 1451–6. Whichever screening test is carried out, the most 12. Barrett-Connor E and Ferrara A. Isolated postchal- important point to appreciate is that all may provide lenge hyperglycemia and the risk of fatal cardiovas- false-negative results (some more than others). Conse- cular disease in older women and men. The Rancho quently, a deteriorating patient will, at the bare min- Bernardo Study. Diabetes Care 1998, 21 (8), 1236–9. imum, require a glucose estimation to be made in a 13. Saydah S, Miret M, Sung J, Varas C, Gause D and fingerprick capillary blood sample. Brancati F. Postchallenge hyperglycemia and mortality in a national sample of U.S. adults. Diabetes Care 2001, 24 (8), 1397–402. References 14. THE DECODE study group on behalf of the European Diabetes Epidemiology Group. Glucose tolerance and 1. National Diabetes Data Group. Classification and di- cardiovascular mortality. Comparison of fasting and agnosis of diabetes mellitus and other categories of 2-h diagnostic criteria. Arch Intern Med 2001, 161, glucose intolerance. Diabetes 1979; 28: 1039–57. 397–404. 3.16 CONCLUSIONS 35

15. Shaw J, Hodge A, de Courten M, Chitson P and Zim- 29. World Health Organization and International Diabetes met P. Isolated post-challenge hyperglycemia con- Federation. Definition and diagnosis of diabetes mel- firmed as a risk factor for mortality. Diabetologia litus and intermediate hyperglycemia: Report of a 1999, 42, 1050–4. WHO/IDF Consultation. World Health Organization, 16. Neely R, Kiwanuka J and Hadden D. Influence of Geneva, 2006. sample type on the interpretation of the oral glucose 30. Croxson SCM, Burden AC, Bodington M and Botha tolerance test for gestational diabetes mellitus. Dia- JL. The prevalence of diabetes in elderly people. betic Medicine 1991, 8, 129–34. Diabetic Medicine 1991, 8 (1), 28–31. 17. Home P. The OGTT: gold that does not shine. Dia- 31. Sinclair AJ, Gadsby R, Penfold S, Croxson SC and betic Medicine 1988, 5 (4), 313–14. Bayer AJ. Prevalence of diabetes in care home resi- 18. Feskens E, Bowles C and Kromhout D. Intra- and dents. Diabetes Care 2001, 24 (6), 1066–8. 32. Harris MI, Hadden WC, Knowler WC and Bennett interindividual variability of glucose tolerance in an PH. Prevalence of diabetes and impaired glucose elderly population. J Clin Epidemiol 1991, 44 (9), tolerance and plasma glucose levels in U.S. population 947–53. aged 20-74 yr. Diabetes 1987, 36 (4), 523–34. 19. Toeller M and Knussmann R. Reproducibility of oral 33. Harris MI, Flegal KM, Cowie CC, Eberhardt MS, glucose tolerance tests with three different loads. Goldstein DE, Little RR, et al. Prevalence of diabetes, Diabetologia 1973, 9 (2), 102–7. impaired fasting glucose, and impaired glucose toler- 20. Mykkanen¨ L, Laakso M, Uusitupa M and Pyorala K. ance in U.S. adults. The Third National Health and Prevalence of diabetes and impaired glucose tolerance Nutrition Examination Survey, 1988-1994. Diabetes in elderly subjects and their association with obesity Care 1998, 21 (4), 518–24. and family history of diabetes. Diabetes Care 1990, 34. Simmons D and Williams D. Diabetes in the elderly: 13 (11), 1099–105. an underdiagnosed condition. Diabetic Medicine 21. Croxson SCM, Price DE, Burden M, Jagger C and 1993, 10 (3), 264–6. Burden AC. The mortality of elderly people with 35. Wang S-L, Pan W-H, Hwu C-M, Ho L-T, Lo C-H, diabetes. Diabetic Medicine 1994, 11 (3), 250–2. Lin S-L, et al. Incidence of NIDDM and the effects 22. Valdes´ S, Botas P, Delgado E, Alvarez F, Cadorniga´ of gender, obesity and hyperinsulinaemia in Taiwan. FD. Population-based incidence of type 2 diabetes Diabetologia 1997, 40, 1431–8. in northern Spain: the Asturias Study. Diabetes Care 36. Haffner S, Lehto S, Ronnemaa¨ T, Pyor¨ al¨ aKand¨ 2007, 30 (9), 2258–63. Laakso M. Mortality from coronary heart disease 23. The Expert Committee on the Diagnosis and Clas- in subjects with type 2 diabetes and in nondiabetic sification of Diabetes Mellitus. Report of the Expert subjects with and without prior myocardial infarction. Committee on the Diagnosis and Classification of Di- N Engl J Med 1998, 339 (4), 229–34. abetes Mellitus. Diabetes Care 1997, 20, 1183–97. 37. British Cardiac Society, British Hypertension Society, 24. McCance D, Hanson R, Charles M, Jacobsson L, Pet- Diabetes UK, HEART UK, Primary Care Cardiovas- titt D, Bennett P, et al. Comparison of tests for gly- cular Society, Stroke Association. JBS 2: Joint British cated haemoglobin and fasting and two hour plasma Societies’ guidelines on prevention of cardiovascular disease in clinical practice. Heart 2005; 91 (Suppl. 5), glucose concentrations as diagnostic methods for di- v1–v52. abetes. British Medical Journal 1994, 308 (6840), 38. Howard-Williams J, Hillson R, Bron A, Awdry P, 1323–8. Mann J and Hockaday T. Retinopathy is associated 25. McHugh D. Letters: Tests for diagnosing diabetes with higher glycaemia in maturity-onset type diabetes. mellitus. Glucose tolerance test is most sensitive. Diabetologia 1984, 27 (2), 198–202. British Medical Journal 1994, 309, 537–8. 39. Hillson R, Hockaday T, Newton D and Pim B. De- 26. Simon K. Letters: Tests for diagnosing diabetes mel- layed diagnosis of non-insulin-dependent diabetes is litus. Glucose tolerance test is most sensitive. British associated with greater metabolic and clinical abnor- Medical Journal 1994, 309, 537–8. mality. Diabetic Medicine 1985, 2 (5), 383–6. 27. Sinclair A. Letters: Tests for diagnosing diabetes mel- 40. Colagiuri S, Cull C, Holman R and the UKPDS litus. Glucose tolerance test is most sensitive. British Group. Are lower fasting plasma glucose levels at di- Medical Journal 1994, 309, 537–8. agnosis of type 2 diabetes associated with improved 28. Brenner H and Gefeller O. Variation of sensitivity, outcomes?: UK prospective diabetes study 61. Dia- specificity, likelihood ratios and predictive values with betes Care 2002, 25 (8), 1410–17. disease prevalence. Statistics in Medicine 1997, 16 41. Testa MA and Simonson DC. Health economic ben- (9), 981–91. efits and quality of life during improved glycemic 36 CH 03 DIABETES IN THE ELDERLY: DIAGNOSIS, TESTING AND SCREENING

control in patients with type 2 diabetes mellitus: a ran- 56. Friderichsen B and Maunsbach M. Glycosuric tests domized, controlled, double-blind trial. JAMA 1998, should not be employed in population screenings for 280 (17), 1490–6. NIDDM. J Public Health Med 1997, 19, 55–60. 42. Stratton IM, Adler AI, Neil HA, Matthews DR, Man- 57. Modan M, Halkin H, Karasik A and Lusky A. Effec- ley SE, Cull CA, et al. Association of glycaemia tiveness of glycosylated hemoglobin, fasting plasma with macrovascular and microvascular complications glucose, and a single post load plasma glucose level of type 2 diabetes (UKPDS 35): prospective obser- in population screening for glucose intolerance. Am J vational study. British Medical Journal 2000, 321 Epidemiol 1984, 119 (3), 4431–44. (7258), 405–12. 58. Wingard DL, Sinsheimer P, Barrett Connor EL and 43. Wachtel TJ, Tetu-Mouradjian LM, Goldman DL, Ellis McPhillips JB. Community-based study of prevalence SE and O’Sullivan PS. Hyperosmolarity and acidosis of NIDDM in older adults. Diabetes Care 1990, 13 in diabetes mellitus: a three-year experience in Rhode (Suppl. 2), 3–8. Island. J Gen Intern Med 1991, 6 (6), 495–502. 59. Wong T, Liew G, Tapp R, Schmidt M, Wang J, 44. Melton LJ, 3rd, Palumbo PJ, Chu CP. Incidence Mitchell P, et al. Relation between fasting glu- of diabetes mellitus by clinical type. Diabetes Care. cose and retinopathy for diagnosis of diabetes: three 1983; 6(1): 75–86. population-based cross-sectional studies. Lancet 2008, 45. Melton LJ, III, Palumbo PJ, Dwyer MS and Chu CP. 371, 736–43. Impact of recent changes in diagnostic criteria on the 60. The DECODE study group on behalf of the European apparent natural history of diabetes mellitus. Am J Diabetes Epidemiology Group. Glucose tolerance and Epidemiol 1983, 117 (5), 559–65. mortality: comparison of WHO and American Di- 46. Butterﬁeld WJ, Keen H and Whichelow MJ. Renal abetes Association diagnostic criteria. Lancet 1999, glucose threshold variations with age. British Medical 354, 617–21. Journal 1967, 4 (5578), 505–7. 61. The DECODE Study Group. Will new diagnostic 47. Singh BM, Jackson DM, Wills R, Davies J and criteria for diabetes mellitus change phenotype of Wise PH. Delayed diagnosis in non-insulin dependent patients with diabetes? Reanalysis of European epi- diabetes mellitus. British Medical Journal 1992, 304 demiological data. British Medical Journal 1998, 317, (6835), 1154–5. 371–5. 48. Jackson DM, Wills R, Davies J, Meadows K, Singh 62. Croxson SCM and Thomas PH. Glucose tolerance BM and Wise PH. Public awareness of the symptoms test results reappraised using recent ADA criteria. of diabetes mellitus. Diabetic Medicine 1991, 8 (10), Practical Diabetes International 1998, 15 (6), 178–80. 971–2. 63. Davies MJ, Ammari F, Sherriff C, Burden ML, Gujral 49. Croxson SCM and Burden AC. Polyuria and poly- J and Burden AC. Screening for Type 2 diabetes dipsia in an elderly population: Its relationship to mellitus in the UK Indo-Asian population. Diabetic previously undiagnosed diabetes. Practical Diabetes Medicine 1999, 16 (2), 131–7. International 1998, 15 (6), 170–2. 64. Piche´ M, Arcand-BosseJ,Despr´ es´ J, Perusse´ L, 50. Croxson S. Diabetes in the elderly. PhD thesis, Uni- Lemieux S and Weisnagel S. What is a normal glu- versity of Leicester, Leicester, 1995. cose value? Differences in indexes of plasma glucose 51. Haavisto M, Mattila K and Rajala S. Blood glucose homeostasis in subjects with normal fasting glucose. and diabetes mellitus in subjects aged 85 years or Diabetes Care 2004, 27 (10), 2470–7. more. Acta Med Scand 1983, 214 (3), 239–44. 65. Balkau B, Shipley M, Jarrett R, Pyor¨ al¨ aK,Py¨ or¨ al¨ aM,¨ 52. Sharp CL, Butterﬁeld WJ and Keen H. Diabetes Forhan A, et al. High blood glucose concentration is Survey in Bedford 1962. Proceedings of the Royal a risk factor for mortality in middle-aged nondiabetic Society of Medicine 1964, 57, 193–202. men. 20-year follow-up in the Whitehall Study, the 53. Forrest R, Jackson C and Yudkin J. Glucose intoler- Paris Prospective Study, and the Helsinki Policemen ance and hypertension in north London: the Isling- Study. Diabetes Care 1998, 21 (3), 360–7. ton Diabetes Survey. Diabetic Medicine 1986, 3 (4), 66. Modan M, Halkin H, Karasik A and Lusky A. Effec- 338–42. tiveness of glycosylated hemoglobin, fasting plasma 54. Wei O and Teece S. Best evidence topic report. Urine glucose, and a single post load plasma glucose level dipsticks in screening for diabetes mellitus. Emerg in population screening for glucose intolerance. Am J Med J 2006, 23 (2), 138. Epidemiol 1984, 119 (3), 431–44. 55. Davies MJ, Williams DR, Metcalfe J and Day JL. 67. Gabir M, Hanson R, Dabelea D, Imperatore G, Community screening for non-insulin-dependent di- Roumain J, Bennett P, et al. The 1997 American abetes mellitus: self-testing for post-prandial glyco- Diabetes Association and 1999 World Health Orga- suria. Q J Med 1993, 86 (10), 677–84. nization criteria for hyperglycemia in the diagnosis 3.16 CONCLUSIONS 37

and prediction of diabetes. Diabetes Care 2000, 23 primary health care. Diabetic Medicine 1993, 10 (2), (8), 1108–12. 167–73. 68. Chang C, Wu J, Lu F, Lee H, Yang Y and Wen M. 81. Bitzen´ P and Schersten´ B. Assessment of laboratory Fasting plasma glucose in screening for diabetes in the methods for detection of unsuspected diabetes in pri- Taiwanese population. Diabetes Care 1998, 21 (11), mary health care. Scand J Prim Health Care 1986, 4 1856–60. (2), 85–95. 69. Dunstan DW Zimmet PZ, Welborn TA, De Courten 82. Croxson SCM, Keir SL and Ibbs L. Admission plasma MP, Cameron AJ, Sicree RA, Dwyer T, Colagiuri S, glucose and diabetes mellitus in elderly admissions to Jolley D, Knuiman M, Atkins R and Shaw JE. The hospital. Diabetic Medicine 1997, 14 (5), 381–5. rising prevalence of diabetes and impaired glucose tol- 83. US Food and Drug Administration. Glucose erance: the Australian Diabetes, Obesity and Lifestyle Meters and Diabetes Management. 2008 [up- Study. Diabetes Care 2002, 25 (5), 829–34. dated 2008; cited 05/03/2008]. Available from: 70. Choi K, Lee J, Kim D, Kim S, Shin D, Kim N, http://www.fda.gov/diabetes/glucose.html#8 et al. Comparison of ADA and WHO criteria for (accessed 1 August 2008). the diagnosis of diabetes in elderly Koreans. Diabetic 84. Atkin S, Dasmahapatra A, Jaker M, Chorost M and Medicine 2002, 19 (10), 853–7. Reddy S. Fingerstick glucose determination in shock. 71. Wahl P, Savage P, Psaty B, Orchard T, Robbins J Ann Intern Med 1991, 114 (12), 1020–4. and Tracy R. Diabetes in older adults: comparison 85. Desachy A, Vuagnat A, Ghazali A, Baudin O, of 1997 American Diabetes Association classification Longuet O, Calvat S, et al. Accuracy of bedside of diabetes mellitus with 1985 WHO classification. glucometry in critically ill patients: Influence of Lancet 1998, 352, 1012–15. clinical characteristics and perfusion index. Mayo 72. Unwin N, Alberti K, Bhopal R, Harland J, Watson Clin Proc 2008, 83, 400–5. W and White M. Comparison of the current WHO 86. Kilpatrick E. Problems in the assessment of glycaemic and new ADA criteria for the diagnosis of diabetes control in diabetes mellitus. Diabetic Medicine 1997, mellitus in three ethnic groups in the UK. Diabetic 14, 819–31. Medicine 1998, 15 (7), 554–7. 87. Jiao Y, Okumiya T, Saibara T, Park K and Sasaki M. 73. The Expert Committee on the Diagnosis and Clas- Abnormally decreased HbA1c can be assessed with sification of Diabetes Mellitus. Report of the Expert erythrocyte creatine in patients with a shortened ery- Committee on the Diagnosis and Classification of Di- throcyte age. Diabetes Care 1998, 21 (10), 1732–5. abetes Mellitus. Diabetes Care 2003, 26 (Suppl. 1), 88. Simon D, Senan C, Garnier P, Saint-Paul M and Papoz S5–S20. L. Epidemiological features of glycated haemoglobin 74. Croxson S and Mostafa S. What is normal Fast- A1c-distribution in a healthy population. The Telecom ing Plasma Glucose? Practical Diabetes International Study. Diabetologia 1989, 32 (12), 864–9. 2008, 25 (5), 209. 89. Altman D and Royston P. The cost of dichotomising 75. WHO Expert Committee. Health of the Elderly. World continuous variables. British Medical Journal 2006, Health Organization, Geneva, 1989. 332, 1080. 76. Croxson S, Bayer A and Sinclair A. Screening for 90. Kilpatrick ES, Dominiczak MH and Small M. The diabetes in care homes. Diabetic Medicine 2005, effects of ageing on glycation and the interpretation 22(Suppl. 2), 100 (P928). of glycaemic control in Type 2 diabetes. Q J Med 77. Simmons D and Williams D. Random blood glucose 1996, 89 (4), 307–12. as a screening test for diabetes in a bi-ethnic popula- 91. Kabadi U. Glycosylation of proteins. Lack of influ- tion. Diabetic Medicine 1994, 11 (9), 830–5. ence of aging. Diabetes Care 1988, 11 (5), 429–32. 78. Engelgau M, Thompson T, Smith P, Herman W, 92. Davidson M, Schriger D, Peters A and Lorber B. Aubert R, Gunter E, et al. Screening for diabetes Revisiting the oral glucose tolerance test criterion for mellitus in adults. The utility of random capillary the diagnosis of diabetes. J Gen Intern Med 2000, 15 blood glucose measurements. Diabetes Care 1995, 18 (8), 551–5. (4), 463–6. 93. Bennett C, Guo M and Dharmage S. HbA1c as a 79. Simmons D, Williams D and Powell M. Prevalence screening tool for detection of Type 2 diabetes: a of diabetes in a predominantly Asian population; systematic review. Diabetic Medicine 2007, 24 (4), preliminary findings of the Coventry diabetes study. 333–43. British Medical Journal 1989, 298, 18–21. 94. Gray CS, Scott JF, French JM, Alberti KGMM and 80. Andersson D, Lundblad E and Svardsudd¨ K. A model O’Connell JE. Prevalence and prediction of unrecog- for early diagnosis of type 2 diabetes mellitus in nised diabetes mellitus and impaired glucose tolerance 38 CH 03 DIABETES IN THE ELDERLY: DIAGNOSIS, TESTING AND SCREENING

following acute stroke. Age Ageing 2004, 33 (1), 107. Kanaya A, Wassel Fyr C, de Rekeneire N, Shorr 71–7. R, Schwartz A, Goodpaster B, et al. Predicting the 95. Fluckiger¨ R, Woodtli T and Berger W. Evaluation of development of diabetes in older adults: the derivation the fructosamine test for the measurement of plasma and validation of a prediction rule. Diabetes Care protein glycation. Diabetologia 1987, 30 (8), 648–52. 2005, 28 (2), 404–8. 96. Croxson SCM, Absalom S and Burden AC. Fruc- 108. Neil H, Thompson A, Thorogood M, Fowler G and tosamine in diabetes screening of the elderly. Annals Mann J. Diabetes in the elderly: the Oxford Commu- of Clinical Biochemistry 1991, 28 (3), 279–82. nity Diabetes Study. Diabetic Medicine 1989, 6 (7), 97. Cefalu W, Bell-Farrow A, Petty M, Izlar C and 608–13. Smith J. Clinical validation of a second-generation 109. Molbak AG, Christau B, Marner B, Borch-Johnsen K fructosamine assay. Clin Chem Lab Med 1991, 37 and Nerup J. Incidence of insulin-dependent diabetes (7), 1252–6. mellitus in age groups over 30 years in Denmark. 98. Cefalu WT, Ettinger WH, Bell-Farrow AD and Rush- Diabetic Medicine 1994, 11 (7), 650–5. ing JT. Serum fructosamine as a screening test for 110. Gale E and Tattersall R. (1990) The new patient: diabetes in the elderly: a pilot study. Journal of the assessment & management. In: R. Tattersall and E. American Geriatrics Society 1993, 41 (10), 1090–4. Gale (eds), Diabetes: Clinical Management. Churchill 99. Ko G, Chan J, Yeung V, Chow C, Tsang L, Li J, Livingstone, Edinburgh, pp. 3–16. et al. Combined use of a fasting plasma glucose 111. Groop L, Tuomi T, Rowley M, Zimmet P and concentration and HbA1c or fructosamine predicts the Mackay I. Latent autoimmune diabetes in adults likelihood of having diabetes in high-risk subjects. (LADA) – more than a name. Diabetologia 2006, 49 Diabetes Care 1998, 21 (8), 1221–5. (9), 1996–8. 100. Lindstrom¨ J and Tuomilehto J. The diabetes risk 112. Gale E. Latent autoimmune diabetes in adults: a score: a practical tool to predict type 2 diabetes risk. guide for the perplexed.. Diabetologia 2005, 48 (11), Diabetes Care 2003, 26 (3), 725–31. 2195–9. 101. Diabetes Prevention Program Research Group. Strate- 113. Umpierrez G, Smiley D and Kitabchi A. Narrative gies to identify adults at high risk for type 2 diabetes: review: ketosis-prone type 2 diabetes mellitus. Ann the Diabetes Prevention Program. Diabetes Care 2005, Intern Med 2006, 144 (5), 350–7. 28 (1), 138–44. 114. Meneilly G and Elahi D. Metabolic alterations in 102. Stern M, Williams K and Haffner S. Identification of middle-aged and elderly lean patients with type 2 persons at high risk for type 2 diabetes mellitus: do diabetes. Diabetes Care 2005, 28 (6), 1498–9. we need the oral glucose tolerance test? Ann Intern 115. Livingstone C and Rampes H. Atypical antipsychotic Med 2002, 136 (8), 575–81. drugs and diabetes. Practical Diabetes International 103. McNeely M, Boyko E, Leonetti D, Kahn S and Fu- 2003, 20 (9), 327–31. jimoto W. Comparison of a clinical model, the oral 116. Balasubramanyam A, Garza G, Rodriguez L, Hampe glucose tolerance test, and fasting glucose for predic- C, Gaur L, Lernmark A, et al. Accuracy and prediction of type 2 diabetes risk in Japanese Americans. tive value of classification schemes for ketosis-prone Diabetes Care 2003, 26 (3), 758–63. diabetes. Diabetes Care 2006, 29 (12), 2575–9. 104. Rathmann W, Martin S, Haastert B, Icks A, Holle 117. Meneilly G, Tildesley H, Elliott T, Palmer J and R, Lowel¨ H, et al. Performance of screening ques- Juneja R. Significance of GAD positivity in elderly tionnaires and risk scores for undiagnosed diabetes; patients with diabetes. Diabetic Medicine 2000, 17 The KORA Survey 2000. Arch Intern Med 2005, 165 (3), 247–51. (436-441), 436. 118. Alberti G, Zimmet P, Shaw J and Grundy SM, for 105. Schmidt M, Duncan B, Bang H, Pankow J, Ballantyne the International Diabetes Federation Task Force on C, Golden S, et al. Identifying individuals at high risk Epidemiology and Prevention. The IDF consensus for diabetes: The Atherosclerosis Risk in Communi- worldwide definition of the METABOLIC SYN- ties study. Diabetes Care 2005, 28 (8), 2013–18. DROME, Brussels, International Diabetes Federation, 106. Heikes KE, Eddy DM, Arondekar B and Schlessinger 2006 (http://www.idf.org/webdata/docs/IDF Meta L. Diabetes Risk Calculator: a simple tool for detect- def final.pdf). ing undiagnosed diabetes and pre-diabetes. Diabetes Care 2008, 31(5), 1040–5. SECTION III Vascular Complications

Leocadio Rodr´ıguez Manas,˜ Cristina Alonso Bouzon and Marta Castro Rodr´ıguez Servicio de Geriatr´ıa, Hospital Universitario de Getafe, Madrid, Spain

perform daily activities. At present, the actual preva- Key messages lence of PAD in the elderly population is unknown, which makes it even more difficult to identify those • Peripheral arterial disease (PAD) is a very cases with gait disorders and/or falls where the under- prevalent disease in the elderly. Often, the lying (or adjuvant) cause is PAD. Accordingly, a high clinical presentation is atypical, and on many index of suspicion has been accorded to PAD when it occasions the only manifestation is a functional is identified as a cause of functional impairment and impairment. falls in old patients. • PAD is a cardiovascular risk factor and a marker Another reason which leads us to ‘ignore’ this dis- of mortality. • The treatment of PAD can improve the patient’s ease is its relatively ‘benign’ evolution. By not compro- quality of life. Hence, an early diagnosis must mising key organs, such as the heart or brain, PAD has be actively sought in the elderly patient with traditionally been regarded as less important than coro- diabetes. nary disease or brain vascular disease. This, in turn, has led us to ignore the serious consequences of the evolution of PAD, including amputation, and the serious 4.1 Introduction prognostic implications of the presence of atherosclerotic lesions in the legs. The mortality rate for PAD at Peripheral arterial disease (PAD) is a condition that 5 years, of approximately 30%, is higher than that of is often ignored, underdiagnosed and poorly treated in many cancers. The majority of these patients will die everyday medical practice. This reality is even greater from vascular complications that do not occur in the among the elderly, in whom the typical symptoms lower limbs, but occur rather in coronary and cerebral of PAD (e.g., pain, claudication) appear in less than territories. The results of recent studies have suggested 50% of cases. Unfortunately, these symptoms may that the risk of a vascular event of any territory is ap- also be mistaken for others that are attributable to proximately twice in patients with PAD as in those conditions that are very prevalent in the elderly, such with coronary or cerebral vascular involvement. as osteoarthritis or neuropathy. For this reason, PAD has in recent years become an Very often – and typically in older patients – the important marker of cardiovascular risk, allowing the only manifestation of PAD is a loss of function, identification of subjects who are at very high cardio- which translates into a progressive loss of autonomy to vascular risk and who require an intensive treatment

Diabetes in Old Age Third Edition Edited by Alan J. Sinclair © 2009 John Wiley & Sons Ltd. ISBN: 978-0-470-06562-4 42 CH 04 PERIPHERAL ARTERIAL DISEASE IN OLD PEOPLE WITH DIABETES of risk factors in order to delay the risk of functional would be even higher if the proportion of apparently impairment. healthy elderly people investigated for its detection were to be increased. The most widely used diagnostic test to detect PAD is the ankle–brachial index, for 4.2 Epidemiology of peripheral which a value <0.9, with a sensitivity >95% and speci- arterial disease ficity close to 100%, is effective in the diagnosis of PAD (see Figure 4.1). In the recently conducted MER- From an epidemiological standpoint, atherothrombosis ITO I study, the prevalence of a low ankle–brachial accounts – either directly or indirectly – for 70% of all index (ABI) was evaluated in older patients with deaths in people aged over 70 years. The factors that metabolic syndrome and the risk factors associated lead to atherosclerotic disease are multiple, and include with its development. The study showed that in those genetic aspects, metabolic diseases, inflammatory dis- patients with metabolic syndrome, factors associated eases, lifestyle and local and systemic conditions of the with a low ABI (<0.9) were age, higher serum creati- vascular system. Among these conditions and diseases nine levels and the presence of proteinuria. After mul- are included the major risk factors, namely older age, tivariate adjustment, only age and active smoking con- smoking, hypertension, dyslipidaemia, type 2 diabetes tinued to be significantly associated with a low ABI [1]. mellitus, physical inactivity and abdominal obesity. In- Typically, patients with PAD have a cardiovascular deed, the INTERHEART study showed that while nine mortality of 2.5%, compared to only 0.5% for healthy of these factors were responsible for 90% of cardiovas- controls. In patients with coronary artery disease, if cular diseases, older age was the only unmodifiable risk PAD is present then the cardiovascular mortality is factor. 25%. In conjunction with PAD serving as an overall While older age is also the main risk factor for PAD, predictor for cardiovascular risk, the ABI may also be other risk factors include smoking, diabetes, hyperten- used as a marker of subclinical functional decline. sion, dyslipidaemia and hyperhomocysteinaemia. • Smoking not only predisposes to develop PAD but also increases its severity. Smoking may also affect 4.3 Pathophysiology the prognosis of revascularization interventions. Atherosclerosis is a process of thickening and stiffening • Diabetes mellitus as a risk factor is both qualitative of the arteries, the basic lesion being the atheroscle- and quantitative. In fact, glycemic control represents rotic plaque. Until recently, the classic concept of one of the most powerful risk factors for illness and atherosclerosis was the mechanical accumulation of for the main consequence of this disease, amputation. lipids and a fibrodegenerative response of the arterial In this regard, amputation is 10-fold more frequent wall by changes in its structure, which in turn caused in diabetic than in non-diabetic patients. a progressive failure of tissue perfusion. Today, these previous ideas have been added to • Although, as a risk factor, hypertension contributes the concept of inflammation, such that atherosclerotic to a lesser extent than either smoking or diabetes disease is regarded as a true multifactorial disease mellitus to the development of PAD, it must also be in which metabolic, inflammatory, hemodynamic and controlled. haemostatic factors are involved, with both local and • With regards to dyslipidaemia, a high total systemic roles. There is no correlation between the his- cholesterol:HDL-cholesterol ratio seems to be the tological phases of the atheroma lesion and clinical best predictor of PAD, and its lowering has been symptoms in the elderly patients. Frequently, the first shown to reduce progression of the disease and in sign appears when the local lesion is advanced; conse- turn the risk of developing critical ischemia. quently, the early active diagnosis of atherosclerosis is very important (see Figure 4.2). • An alteration in the metabolism of homocysteine is It is important, therefore, that the presence of an important risk factor for atherosclerosis; however, atherosclerosis is detected at an early stage. To that its role is most likely age-dependent, as it is more end, it is essential to identify any physiological dis- relevant in young people. orders associated with endothelial dysfunction and its While the prevalence of PAD in the elderly (considered progression toward atherothrombosis, long before these as aged >70 years) is generally 15–20%, this figure become visible obstructive lesions. This highlights the 4.3 PATHOPHYSIOLOGY 43

Formula Right ABI = highest right ankle pressure/highest arm Pressure at pressure right or left Left ABI = highest left ankle pressure/highest arm arm

MATERIAL NEEDED Portable Doppler Sphygmomanometer 8 mHz transducer

Results of index >0.90 - Normal <0.90 - obstruction Pressure at posterior <0.40 - severe obstruction tibial and dorsalis pedis arteries in right and left ankle

Figure 4.1 The ankle–brachial index.

Atherosclerosis: a silent, widespread and gradual process

Break/cleft/ Normal Deposit of Fibrou Atherosclerotic of the plaque & Artery Cholesterol splaque plaque thrombosis ANGOR

AMI

Age, Sex, Genetic factors PAD High Cholesterol Smoking STROKE HTA Diabetes Mellitus Obesity and sedentary lifestyle

Cardiovascular risk factors

Figure 4.2 The evolution of atherosclerosis. 44 CH 04 PERIPHERAL ARTERIAL DISEASE IN OLD PEOPLE WITH DIABETES importance of the early treatment of vascular risk exercise, but is sufficient when the patient is resting. factors. It must be remembered that vascular disease Such functional ischemia translates clinically as inter- is not simply a local process within a concrete plaque, mittent claudication. but rather is a widespread process that affects the entire Critical ischemia occurs when the flow is insufficient vascular tree. This concept has therapeutic implications even at rest, and appears as pain and trophic lesions in because it forces us to raise an integrated treatment. the extremities. In this situation there is a need to inter- Another important aspect of the pathophysiology vene to restore an adequate blood flow, so as to avoid of atherosclerosis (more specifically of endothelial the risk of amputation. However, the symptoms will dysfunction) in the elderly is which of the processes depend largely on the number of affected territories and that occur at the endothelium can be attributed to the level of physical activity that the subject demands. physiological aging, and which to the presence of Glyco-oxidation contributes to the development of other cardiovascular risk factors. Some of the mech- atherosclerosis in the below-the-knee peripheral artery anisms involved in the development of endothelial tree in type 2 diabetes. Advanced glycosylation end dysfunction are in fact shared by both aging and car- product (AGE) levels are increased in type 2 diabetic diovascular risk factors (e.g., diabetes, hypertension, patients with PAD, as compared to levels in diabetic hypercholesterolemia) [2]. A knowledge of these patients without PAD and in control subjects. More issues has deep practical therapeutic implications, as precisely, among the components of AGEs, pentosi- they may serve as the main constraint on the intensity dine appears to be strongly associated with the periph- of cardiovascular risk factor treatment. eral artery status of diabetic patients. In addition, lipid Peripheral arterial insufficiency occurs when the oxidation, as estimated by the serum levels of malondi- blood flow reaching the limbs is insufficient to fulfil aldehyde (MDA), is associated with diabetic peripheral the metabolic needs of the tissues. This often results angiopathy. In contrast, both total reactive antioxidant from the presence of an occlusive arterial disease, this potentials (TRAP) and vitamin E levels, as expressions being the underlying disease process of atherosclerosis of a defence mechanism against glycolipid oxidation, and affecting primarily – though not exclusively – the are lower in type 2 diabetic patients with PAD than in vascularization of the lower limbs. those without PAD and in healthy subjects [3]. Although atherosclerosis – and in particular the for- However, the presence of several cardiovascular risk mation of atherosclerotic plaque – is an universal pro- factors, which act in synergistic manner, is an im- cess, it does show certain pathophysiological differ- portant factor in progression of the disease and am- ences, depending on the anatomic location in ques- putation. However, not all cardiovascular risk factors tion. Atherosclerotic plaques located in high-risk lower contribute equally: for example, diabetes mellitus in- limbs are very fibrous and strictures and, when asso- creases the risk of critical ischemia fourfold and smok- ciated with a hypercoagulable state, this gives rise to ing threefold, while an ABI <0.5 increases such risk an acute event. By contrast, at level of the coronary 2.5-fold. In addition to these factors, in the case of am- arteries, the atherosclerotic plaque consists of a large putation, there are other independent risk factors for extracellular lipid core and a large number of foam amputation, notably sensory neuropathy, PAD, previ- cells, coated with a thin cover that is susceptible to ous minor amputations and the use of insulin [4]. breakage; such breakage is the ultimate cause of any In the Wisconsin study, which followed a cohort that acute event that might happen. In any event, common included elderly patients for 14 years, male gender, to all these injuries there is a consequence of an imbal- high levels of HbA1c, high pulse pressure and severe ance between the needs of the tissues and blood flux. retinopathy were associated with a greater power to a If this mismatch occurs suddenly, as in a thrombotic need for amputation, while regular aspirin consumption event, then it will lead to acute ischemia. However, was protective [5]. if the establishment of a stenosis is gradual, allowing In addition, other factors that are very prevalent in the development of a collateral circulation, and/or a the elderly (e.g., physical disability, loss of vision or a metabolic adaptation of the muscle mass involved and shortage of social resources) act as facilitators of the the use of non-ischemic muscle groups, then ischemia amputation. However, in the pathogenesis of diabetic may persist as a chronic state (see Figure 4.2). amputations in the elderly, coexisting involvement of From the pathophysiological point of view, it is the peripheral nervous system, microvascular damage possible to refer to functional ischemia when the and infection are the most important concurrent fac- blood flow is insufficient for any demand that involves tors. Peripheral neuropathy diminishes pain perception, 4.4 CLINICAL PRESENTATION 45 placing the skin at risk for suffering from continu- some studies [7, 8] have demonstrated how intermit- ous damage and muscle atrophy. As a consequence, tent claudication may underestimate the prevalence of changes in the points of support occur, downloading PAD. The reasons for such asymptomatic presentation the pressure on areas not prepared for such change. might be the development of collateral arteries, a mus- As a result the metatarsal heads may suffer ischemic cular adjustment to the ischemia, and the employment necrosis which, in the presence of ulcers, might facil- of muscular groups least affected. Added to this, the itate the development of osteomyelitis. Sensory neu- presence of comorbidity and functional impairment in ropathy hinders the perception of pain as a symptom some elderly patients do not allow them to perform a of alarm, thereby facilitating the emergence of pressure sufficiently active life to provoke intermittent claudica- ulcers. The autonomic neuropathy facilitates the open- tion. On the other hand, chronic ischemia can modify ing of artery–vein shunts and causes skin hydration to the muscle function. In a sub-study from the Women’s become difficult. Health and Aging Study [9], those patients with a low The involvement of the microcirculation has been ABI but without claudication were characterized by a widely discussed. Although the basement membrane is slow walking, a longer time to arise from a seated posi- seen to become thickened, this does not seem to be tion than normal, a poor standing balance score, and a clinically significant in the absence of peripheral and shorter weekly walking distance, even after making ad- autonomic neuropathy. Other factors, which are related justments for age, gender, race, cigarette smoking and both directly and indirectly to vascular injury, cooper- comorbidities. Similar results were found elsewhere ate in the development of clinically apparent damage: [10, 11], highlighting the importance of functional im- the ischemia causes pain, especially in patients with pairment as a frequent means of clinical presentation high blood glucose levels, and difficulty may be en- of PAD in older people. Furthermore, those patients countered in the healing of existing injuries while the with a low ABI but no claudication do not enjoy a sterilization of infected lesions may be delayed. Other benign functional [9, 12] and cardiovascular [13, 14] mechanisms that can hinder healing include AGEs or a prognosis. In fact, it is important to note that there is no zinc shortfall in relation to its increased renal elimina- relationship between the injury severity and the clini- tion in patients with poor glycemic control. It is quite cal presentation of the disease, with the severity of the possible that this mechanism should be enhanced in vascular injuries being the most important prognostic the elderly. The ischemia that occurs due to a poor cir- factor. culation also prevents the antibiotic from targeting the infected ulcers. Finally, other mechanisms associated with hypergly- 4.4.2 Claudication cemia may participate in the pathogenesis of the dia- Claudication is defined as fatigue, discomfort or pain betic foot, but this is more controversial. It is especially that occurs in specific limb muscle groups during ef- the case for the decline in chemotaxis, phagocytosis fort, as a result of exercise-induced ischemia [15]. and bacterial lysis secondary to hyperglycemia [6]. These symptoms are absent at rest. It is important to Among elderly patients, while amputation remains determine a differential diagnosis from other causes of the most dramatic complication, as it creates not only ischemia (e.g., emboli, Berger’s disease, other arter- disability but also psychological damage, it must not itides), and in particular claudication must be distin- be forgotten that on many occasions the cause of func- guished from other illnesses that cause exertional leg tional decline – notably in those with sarcopenia – is pain – so-called ‘pseudoclaudication’ – as a result per- that of PAD. haps of lumbar disease and spinal stenosis, osteoarthritis, severe venous obstructive disease and peripheral 4.4 Clinical presentation neuropathy. The anatomic site of an arterial stenosis has been associated frequently with specific leg symptoms. For 4.4.1 Asymptomatic example, obstructions in the femoral and popliteal Although intermittent claudication (see Section 4.4.2) arteries are associated with calf pain, while affectation is the most common symptom in patients with PAD, in the tibial arteries may produce calf pain, foot pain or the majority of individuals with this pathology do not numbness. Occlusive disease in the iliac arteries could experience this typical limb ischemic symptom. Rather, produce hip, buttock, thigh pain and calf pain. 46 CH 04 PERIPHERAL ARTERIAL DISEASE IN OLD PEOPLE WITH DIABETES

Table 4.1 Fontaine’s stages of the severity of clinical Although generally the progression of PAD from symptoms of patients with PAD. Values in parentheses asymptomatic and intermittent claudication to criti- indicate walking distances. cal limb ischemia will occur gradually, occasionally a Fontaine stage Clinical finding rapid or sudden decrease in limb perfusion will threaten tissue viability. Pain, paralysis, paresthesias, pulseless- I Asymptomatic ness and pallor are the five ‘P’s that suggest this syn- IIa Mild claudication (>150 m) IIb Moderate-severe claudication (<150 m) drome. In this case it is imperative that the patient is III Ischemic rest pain evaluated by a vascular surgeon as an emergency case. IV Ulceration or gangrene 4.5 Diagnostic methods The severity of the clinical symptoms allows patients 4.5.1 Anamnesis and physical assessment with PAD to be categorized (see Table 4.1). This is not only especially useful for facilitating communication The diagnosis of PAD is based mainly on a clinical between different specialists, but also has important evaluation, and the medical history. Very often, a pa- therapeutic value. Although intermittent claudication tient will minimize the symptoms, attributing them to is often poorly correlated with the actual stenosis, the normal aging; consequently, an active search for in- symptoms, their repercussions on the quality of life and termittent claudication or any atypical presentation of the potential benefits with different treatment strate- PAD must be carried out, thereby distinguishing any gies, are each keys to decide between revascularization symptoms from those of a non-vascular cause (pseu- and conservative treatment. doclaudication) [15, 17]. A careful history, including a comprehensive geriatric assessment, may allow the discovery of functional impairment at clinical presen- 4.4.3 Critical limb ischemia tation. A physical assessment should include an examina- Critical limb ischemia is defined as limb pain which tion with the patient’s shoes and socks removed, paying occurs at rest, or as the immediate limb loss caused by special attention to pulses (femoral, popliteal, posterior a severe compromise of blood flow to the affected ex- tibial and pedal), bruits, hair loss, skin colour, temper- tremity [15]. This pain, which may be sharp in nature, ature, ulcers and trophic skin changes [18]. occurs typically at rest, especially when the patient is The clinical guidelines for type 2 diabetes mellitus supine or when the leg is held in an elevated posi- (European Diabetes Working Party for Older People tion. Occasionally, narcotics may be needed to control 2004) [19] suggest that a comprehensive geriatric as- the pain, but these may often cause severe side ef- sessment should be routine in older people with type 2 fects in the elderly. Consequently, such drugs should diabetes, both at the initial diagnosis and thereafter at be avoided whenever possible, but if unavoidable they regular intervals. Further recommendations include (at should be used with extreme caution. These patients minimum) an annual inspection of the feet by a health may also suffer trophic skin changes or tissue loss, care professional, including vascular and neurological ulcers or gangrene. Ischemic ulcers are very painful examinations, even in the absence of any symptoms. (unless they are associated with neuropathy), and have Other physical measures, such as a slow walking irregular margins, no pulses, are dry, and occur fre- speed, a longer time to rise from a seated position, quently on the toes with a cold, pale or cyanotic foot. or a poor standing balance score, have been related to These ulcers are usually associated with infection in the presence of subclinical disease [9]. At present, no the surrounding tissues. evidence is available of the sensitivity and prognostic Critical limb ischemia may represent the initial pre- values of these latter symptoms/signs within the usual sentation of a lower-extremity PAD, and occur more clinical setting. frequently among elderly diabetic patients. This situation arises because the arterial disease develops es- 4.5.2 Vascular diagnostic techniques pecially on small arteries [16]; there is also often an These tests allow the objective establishment of a important comorbidity such that the intermittent clau- diagnosis of PAD, for the severity of the disease to dication is suppressed. be quantified, the stenosis to be localized, a plan 4.6 TREATMENT 47 of treatment to be organized, and the progression of and fissuring, as well as daily inspections and cleansing disease or its response to treatment to be determined. by the patient and chiropodist, are necessary measures The ABI is a quick and cost-effective method to reduce the risk of skin ulceration, necrosis and am- for providing sufficient information for the screen- putation. These measures are recommended based on ing, diagnosis and follow-up. In order to calculate the results of studies in which their effects in diabetic the ABI, the systolic blood pressures are determined patients have been analyzed [25, 26]; however, no in- in both arms and both ankles, using a hand-held vestigations appear to have been carried out analyzing Doppler instrument (see Figure 4.1). Although the the impact of such measures in elderly diabetic patients ABI has been validated against the ‘gold standard’ with PAD. of lower-extremity contrast angiography [20, 21], in patients with non-comprehensible arteries, such as 4.6.2 Cardiovascular risk reduction long-term diabetics and very old patients, it may not be sufficiently accurate, raising the probability of obtain- The treatment of patients with PAD requires that each ing false-negative results. In these cases, other alterna- modifiable risk factor associated with the develop- tive non-invasive diagnostic tests (e.g., the toe–brachial ment and evolution of the condition be addressed, index, exercise ABI test, pulse volume recording) including cigarette smoking, diabetes mellitus, seden- should be performed [15]. tary lifestyle, dyslipidaemia, hyperhomocystinaemia An altered ABI has been associated with systemic and hypertension. To date, no conclusive evidence has atherosclerotic disease, total and cardiovascular mortal- been reported regarding the relationship between the ity [22, 23] and functional impairment [9]. Today, the control of risk factors and the prognosis of PAD. Nev- American Diabetes Association (ADA) suggests that ertheless, the need to control cardiovascular risk factors the ABI must be performed by the general practitioner in the manifestation of atherosclerosis has been well for all patients with diabetes who are aged ≥50 years established, as has the association between PAD and (i.e., in every old patient), in diabetic individuals aged systemic atherosclerosis. Consequently, the screening <50 years who have other atherosclerosis risk factors, and treatment of cardiovascular risk factors should be and in those patients who have been diabetic for at regarded as a priority in all patients with PAD, inde- least 10 years [24]. pendent of their clinical manifestations. Other possible diagnostic techniques include mag- No evidence has yet been provided to suggest the netic resonance angiography and computed tomo- preferred control level of cardiovascular risk factors in graphic angiography, both of which can be used to elderly diabetic patients with PAD. Hence, the treat- identify the anatomic location and degree of a steno- ment goals are similar to those for diabetic elderly sis. These methods are valuable when selecting pa- patients [15]. Antiplatelet therapy with aspirin, using daily doses tients, notably those for endovascular treatment. The of 75 to 325 mg, reduces the risk of vascular death, diagnostic performance of both imaging procedures is myocardial infarction and stroke in patients with PAD quite similar, but computed tomographic angiography [27]. Although clopidogrel appears to be more effective is preferred when magnetic resonance is contraindi- than aspirin in preventing ischemic events in individu- cated. Today, contrast angiography is regarded as the als with symptomatic PAD [28], the size effect does not ‘gold standard’, and is the definitive method before allow any broad recommendation to be made regard- revascularization procedures. However, it may be asso- ing its use instead of aspirin. Thus, the more expensive ciated with a higher risk of medical complications (e.g., thienopyridines (ticlopidine and clopidogrel) may be bleeding, infection, contrast allergy) than non-invasive considered as alternatives to aspirin when patients are techniques and should be performed only in selected unable to tolerate the latter. Currently available data (surgical) patients [15]. have not indicated any advantage of dual antiplatelet therapy over single-agent therapy. 4.6 Treatment 4.6.3 Intermittent claudication 4.6.1 Foot care In addition to the previously described approaches The use of appropriate footwear to avoid pressure in- to treat intermittent claudication, it is important also juries, the use of moisturizing cream to prevent dryness to consider exercise and rehabilitation, as well as 48 CH 04 PERIPHERAL ARTERIAL DISEASE IN OLD PEOPLE WITH DIABETES pharmacological and endovascular treatments focused 3. Lapolla A, Piarulli F, Sartore G, Ceriello A and Ragazzi on specific manifestations of the condition [15]. There E. Advanced glycation end products and antioxidant is strong evidence supporting the beneficial effects status in type 2 diabetic patients with and without pe- of exercise [29, 30], with regular walking, within a ripheral artery disease. Diabetes Care 2007; 30: 670–6. supervised claudication exercise program, having been 4. Adler AJ, Boyko EJ, Ahroni JH and Smith DG. Lower shown to improve walking time free of pain by an extremity amputation in diabetes. The independent effects of peripheral vascular disease, sensory neuropathy average of 150% (range: 74 to 230%). Such a program and foot ulcers. Diabetes Care 1999; 22: 1029–35. was based on certain specifications, with the patient 5. Moss SE, Klein R and Klein BEK. The 14-year in- walking at close to the maximum tolerable pain, for cidence of lower-extremity amputations in a diabetic more than 30 min per session, three times each week population. Diabetes Care 1999; 22: 951–9. for more than six months. 6. Rodr´ıguez-Manas˜ L and Monereo Meg´ıas S. (2002) Cilostazol is a phosphodiesterase type 3 inhibitor El anciano con diabetes. Sociedad Espanola˜ de with vasodilatory and antiplatelet properties. Following Medicina Geriatrica´ (Spanish Society of Geriatric cilostazol treatment (100 mg, twice daily), the walking Medicine)/Sociedad Espanola˜ de Endocrinolog´ıa y distance was increased by approximately 50% com- Nutricion.´ Villanueva, Madrid. pared to placebo, after 3–6 months of therapy [31]. 7. Meijer WT, Hoes AW, Rutgers D, Bots ML, Hofman Unfortunately, cilostazol is contraindicated in patients A and Grobbee DE. Peripheral arterial disease in the with heart failure, which limits its use in old people elderly: the Rotterdam Study. Arterioscler Thromb Vasc with diabetes, where heart failure is very common. Biol 1998, 18: 185–92. Pentoxifylline has been considered a second-line ther- 8. Criqui MH, Fronek A, Barret-Connor E, Klauber MR, Gabriel S and Goodman D. The prevalence of periph- apy to improve walking distances, but its efficacy is eral arterial disease in a defined population. Circulation not well established; indeed, this is also the case for 1985, 71: 510–15. other pharmacological agents, including oral vasodila- 9. McDermott MM, Ferrucci L, Simonsick EM, Balfour tor prostaglandins, vitamin E and ginkgo biloba. J, Fried L, Ling S, Gibson D and Guralnik JM. The Endovascular or surgical revascularization therapy ankle brachial index and change in lower extremity is reserved for those patients in whom the functional functioning over time: the Women’s Health and Aging capacity is compromised only by claudication (not for Study. J Am Geriatr Soc 2002; 50: 238–46. other comorbidities), for those who have not responded 10. McDermott MM, Greenland P, Liu K, Guralnik JM, to exercise and pharmacotherapy, and for those in Criqui MH, Dolan NC, Chan C, Celic L, Pearce WH, whom the risk–benefit ratio with revascularization is Schneider JR, Sharma L, Clark E, Gibson D and Martin favourable [15]. These patients, and those with criti- GJ. Leg symptoms in peripheral arterial disease: associ- cal and acute limb ischemia, should be referred to a ated clinical characteristics and functional impairment. vascular surgeon. JAMA 2001, 286: 1599–606. Recently, the role of stem or progenitor cells 11. Doland NC, Liu K, Criqui MH, Greenland P, Gural- nik JM, Chan C, Schneider JR, Mandapat AL, Martin in vascular disease, including atherosclerosis and G and McDermott MM. Peripheral artery disease, di- post-angioplasty restenosis, has been demonstrated abetes, and reduced lower extremity functioning. Dia- [32–35]. Although such studies utilized animal betes Care 2002, 25: 113–20. models, this line of investigation clearly constitutes a 12. McDermott MM, Liu K, Greenland P, Guralnik JM, new approach in the treatment of PAD. Criqui MH, Chan C, Pearce WH, Schneider JR, Fer- rucci L, Celic L, Taylor LM, Vonesh E, Martin GJ and Clark E. Functional decline in peripheral arterial dis- References ease: associations with the ankle brachial index and leg symptoms. JAMA 2004, 292: 453–61. 1. Suarez C, Manzano L, Mostaza J, Cairols M and Palma 13. Long TH, Criqui MH, Vasilevskis EE, Denenberg JO, JC. Prevalence of peripheral artery disease evaluated Klauber MR and Fronek A. The correlation between by ankle brachial index in patients with metabolic the severity of peripheral arterial disease and carotid syndrome. MERITO I study. Rev Clin Esp. 2007; 207: occlusive disease. Vasc Med 1999; 4: 135–42. 228–33. 14. Criqui MH and Denenberg JO. The generalized nature 2. Serrano Hernando FJ and Mart´ın Conejero A. Pe- of atherosclerosis: how peripheral arterial disease may ripheral artery disease: pathophysiology, diagnosis and predict adverse events from coronary artery disease. treatment. Rev Esp Cardiol 2007; 60: 969–82. Vasc Med 1998, 3: 241–5. 4.6 TREATMENT 49

15. Hirsch AT, Haskal ZJ, Hertzer NR, et al. ACC/AHA 26. Reiber GE, Smith DG, Wallace C, Sullivan K, Hayes 2005 Practice Guidelines for the management of pa- S, Vath C, Maciejewski ML, Yu O, Heagerty PJ and tients with peripheral arterial disease (lower extremity, LeMaster J. Effect of therapeutic footwear on foot renal, mesenteric, and abdominal aortic): a collabora- reulceration in patients with diabetes: a randomized tive report from the American Association for Vascu- controlled trial. JAMA 2002, 287: 2552–8. lar Surgery/Society for Vascular Surgery. Circulation 27. Antithrombotic Trialists’ Collaboration. Collaborative 2006, 113: e463–e654. metaanalysis of randomised trials of antiplatelet ther- 16. Aboyans V, Criqui MH, Denenberg JO, Knoke JD, apy for prevention of death, myocardial infarction and Ridker PM and Fronek A. Risk factors for progression stroke in high risk patients. Br Med J 2002, 324: of peripheral arterial disease in large and small vessels. 71–86. Circulation 2006, 113: 2623–9. 28. CAPRIE Steering Committee. A randomised, blinded, 17. Schmieder FA and Comerota AJ. Intermittent claudi- trial of clopidogrel versus aspirin in patients at risk cation: magnitude of the problem, patient evaluation, of ischaemic events (CAPRIE). Lancet 1996, 348: and therapeutic strategies. Am J Cardiol 2001, 87: 1329–39. 3D–13D. 29. Leng GC, Fowler B and Ernst E. Exercise for intermit- 18. White, C. Intermittent claudication. N Engl J Med 2007; tent claudication. Cochrane Database Syst Rev, 2000: 356: 1241–50. CD000990 19. European Diabetes Working Party for Older People. 30. Bendermacher BL, Willigendael EM, Teijink JA (2004) Clinical guidelines for Type 2 Diabetes Mel- and Prins MH. Supervised exercise therapy versus litus. www.eugms.org/index.php?pid=30 (accessed 6 non-supervised exercise therapy for intermittent December 2008). claudication. Cochrane Database Syst Rev, 2006: 20. Lijmer JG, Hunink MG, van den Dungen JJ, Loonstra CD005263. J and Smit AJ. ROC analysis of noninvasive tests for 31. Strandness DE, Jr, Dalman RL, Panian S, Rendell peripheral arterial disease. Ultrasound Med Biol 1996, MS, Comp PC, Zhang P and Forbes WP. Effect of 22: 391–8. cilostazol in patients with intermittent claudication: 21. Feigelson HS, Criqui MH, Fronek A, Langer RD and a randomized, double-blind, placebo-controlled study. Molgaard CA. Screening for peripheral arterial disease: Vasc Endovascular Surg 2002, 36: 83–91. the sensitivity, speciﬁcity and predictive value of non- 32. Nevskaya T, Ananieva L, Bykovskaia S, Eremin I, invasive tests in a deﬁned population. Am J Epidemiol Karandashov E, Khrennikov J, Mach E, Zaprjagaeva 1994, 140: 526–34. M, Guseva N and Nassonov E. Autologous progenitor 22. Newman AB, Siscovick DS, Manolio TA, Polak J, cell implantation as a novel therapeutic intervention for Fried LP, Borhani NO and Wolfson SK. Ankle-arm ischaemic digits in systemic sclerosis. Rheumatology index as a marker of atherosclerosis in the Cardio- 2009; 48 (1): 61–4. vascular Health Study (CHS) Collaborative Research 33. Werner N and Nickenig G. Clinical and therapeutical Group. Circulation 1993, 88: 837–45. implications of EPC biology in atherosclerosis. JCell 23. Newman AB, Sutton-Tyrrel K, Vogt MT and Kuller Mol Med. 2006; 10 (2): 318–32. LH. Morbidity and mortality in hypertensive adults 34. Versari D, Lerman LO and Lerman A. The importance with a low ankle/arm blood pressure index. JAMA of reendothelialization after arterial injury. Curr Pharm 1993, 270: 487–9. Des. 2007; 13 (17): 1811–24. 24. American Diabetes Association. Peripheral arterial dis- 35. Kiernan TJ, Yan BP, Cruz-Gonzalez I, Cubeddu RJ, ease in people with diabetes. Diabetes Care 2003, 26: Caldera A, Kiernan GD and Gupta V. Pharmacolog- 3333–41. ical and cellular therapies to prevent restenosis af- 25. Plank J, Haas W, Rakovac I, Gorzer¨ E, Sommer R, ter percutaneous transluminal angioplasty and stenting. Siebenhofer A and Pieber TR. Evaluation of the impact Cardiovasc Hematol Agents Med Chem. 2008; 6 (2): of chiropodist care in the secondary prevention of foot 116–24. ulcerations in diabetic subjects. Diabetes Care 2003, 26: 1691–5. 5 Coronary Heart Disease

Ahmed H. Abdelhaﬁz Department of Elderly Medicine, Rotherham General Hospital, Rotherham, UK

develop coronary heart disease (CHD) they have at Key messages least twice excess risk of morbidity and worse cardiovascular outcomes compared to non-diabetics. More- • Cardiovascular disease is the single most impor- over, myocardial ischemia due to coronary atheroscle- tant cause of death in patients aged 70 years and rosis is commonly silent in diabetics. As a result, CHD over, and diabetes is an independent risk factor is often present before ischemic symptoms occur. Hy- for cardiovascular disease. perglycemia, insulin resistance, hyperinsulinaemia and • In older people with diabetes, signiﬁcant evi- visceral obesity, in addition to ‘traditional’ risk factors, dence of coronary heart disease may be present, are the major contributors to CVD in diabetics. In this with few or no symptoms. • Optimizing cardiovascular care requires a mul- chapter we review the major risk factors for CHD, MI tifaceted approach that targets risk factor iden- and heart failure (HF) in diabetes, with particular em- tiﬁcation, hypertension, dyslipidaemia, the hy- phasis on evidence relevant to older people (as much percoagulable state, as well as glucose control. as available). Because type 2 diabetes affects the vast majority of older people diagnosed with diabetes, it will be the primary focus of the chapter. 5.1 Introduction 5.2 Effect of ageing and diabetes on Diabetes mellitus has been recognized as an indepen- the cardiovascular system dent major cardiovascular risk factor since the publication of the Framingham study in 1979 [1]. In spite Both, ageing and diabetes have profound effects on the of various known metabolic and microvascular com- cardiovascular system structure and function. Indeed, plications of diabetes, cardiovascular disease (CVD) such related changes to the cardiovascular system are remains the most common cause of death in all age themselves increasingly recognized as risk factors for groups. Diabetes itself, in the absence of associated CVD. CVD, constitutes a risk similar to those of non-diabetic individuals with previous history of myocardial infarc- 5.2.1 The effect of ageing tion (MI) [2]. In that sense, diabetes is considered as a ‘coronary risk equivalent’, and any prevention in Vascular ageing contributes to the age-dependent rise this population should be considered as a secondary in hypertension and atherosclerotic disease. With in- prevention. In fact, it may be appropriate to say that creasing age, the intima of the arterial wall becomes diabetes is a CVD. On top of that, when diabetics less smooth, with an increased deposition of lipid, cal-

Men 10 Women Annual Incidence per 1000 Annual 0 35-44 45-54 55-64 65-74 75-84 Age

Figure 5.1 Incidence of coronary heart disease by age in men (light bars) and women (dark bars). Reproduced with permission from Ref. [12].

cium and connective tissue. In the media, there is an and endothelial dysfunction in apparently healthy older increased deposition of elastic fibres and smooth mus- people lead to increases in the systolic and pulse pres- cle cells, which in turn leads to a stiffening of the sures that were formerly thought to be part of normal vascular wall and a loss of compliance of the aorta ageing. This vascular ageing precedes and predicts a and major arteries [3]. This loss of compliance plays higher risk for the development of clinical disease; in- a central pathophysiological role in systolic hyperten- deed, what is now referred to as vascular disease could sion in older people. Isolated systolic hypertension and be regarded as the vascular ageing–vascular disease in- a wide pulse pressure are a result of reduced compli- teraction [12]. This results in a steep increase in the ance, and are a powerful determinant of cardiovascular prevalence and incidence of CHD by increasing age risk [4]. The elastic recoil of the central arteries in di- (Figure 5.1) [12]. astole is important for coronary perfusion, as a loss of such elasticity impairs the coronary blood flow and 5.2.2 The effect of diabetes may also contribute to the development of CHD. Both, arterial function and vascular ageing, could be pro- Arterial endothelium normally produces vasoactive grammed during fetal life or influenced by adverse substances such as nitric oxide (vasodilator) and en- growth patterns in early postnatal life [5]. However, dothelin (vasoconstrictor and procoagulant). In dia- interindividual variations in the risk of occurrence of betes, when endothelin production increases and nitric ageing changes, and the development of CVD may be oxide production decreases; this favours a procoagulant genetic in origin [6, 7]. The responsiveness of arte- state and promotes vascular smooth muscle growth, rioles to the catecholamine alpha-adrenergic receptor causing in turn an increased risk of cardiovascular also decreases with ageing, and this prevents an ad- events [13]. The formation of advanced glycosylation equate increase in systemic vascular resistance from end products also increases in diabetes, and this leads occurring in older individuals during orthostasis [8], to structural wall changes and endothelial dysfunction making them prone to drug-induced orthostatic hy- [14, 15]. The increased risk of CVD in diabetes is potension [9]. not fully explained by the traditional risk factors, and At the level of the myocardium, a prolonged expo- there is some evidence to suggest that abnormalities sure to a high systolic pressure leads to an increased of insulin-like growth factor-1 and one of its binding myocyte turnover, with subsequent hypertrophy and in- proteins, insulin-like growth factor binding protein-1, terstitial fibrosis. This results in a stiff, non-compliant occur in insulin-resistant states and may be significant myocardium, which in turn leads to diastolic dys- factors in the pathophysiology of CVD [16]. function and an impaired early diastolic filling of the At the level of the myocardium, diabetes is asso- ventricles [10]. Left ventricular hypertrophy has been ciated with an abnormal left ventricular structure and shown to be associated with an increased risk of CVD function in older people; typically, the left ventricu- [11]. The increased large artery thickening, stiffness lar mass increases in diabetes. In the Cardiovascular 5.4 CARDIOVASCULAR RISKS 53 Health Study, among a cohort of 5201 men and women 5.3 Epidemiology aged ≥65 years, the echocardiographically measured ventricular septal and left posterior wall thicknesses Today, as the prevalence of type 2 diabetes contin- were greater in diabetic than in non-diabetic subjects, ues to increase rapidly in the developed world, due showing a significant linear trend with increased du- to the ageing of the population and an increased fre- = ration of diabetes (p 0.025 for ventricular septal quency of obesity, the prevalence of associated CVD = thickness; p 0.002 for posterior wall thickness). An is also increasing [21]. This is likely to lead to an increased wall thickness of the ventricular septum or epidemic of CHD over the next decades. Currently, the left posterior wall was not associated with preva- CVD is the most common cause of death in diabetic lent CHD in the cohort. After adjusting for body patients, affecting between 65% and 80%, compared weight, blood pressure, heart rate and prevalent coro- to only one-third of all deaths in the general popu- nary or cerebrovascular disease, diabetes remained lation [22]. The prevalence of CHD is approximately an independent predictor of increased left ventricu- 80% of elderly Americans with type 2 diabetes [23]. lar mass among men and women (174.2 g in dia- Silent or asymptomatic CHD is also highly prevalent betic men versus 169.8 g in normal men; 138.2 g ver- among diabetics, with autopsy studies having reported = sus 134.0 g, respectively, for women; p 0.043 for a prevalence of CHD in diabetics but without ante- both genders combined). This association between di- mortem evidence of clinical CHD, ranging from 50% abetes and left ventricular mass appeared to be both to 75% [24, 25]. Screening for CHD in diabetic pa- duration- and severity-dependent [17]. Diabetes also tients will not alter risk factor management, because impairs cardiac diastolic function, leading to a myo- these patients are considered at high risk on the ba- pathic state known as diabetic cardiomyopathy.This sis of diabetes alone [26]. However, screening may be involves a prolongation of contraction and relaxation, useful in high-risk patients in whom revascularization as well as a slowing in relaxation velocity [18]. Po- therapy will be indicated. This is particularly important tential abnormalities underlying this cardiomyopathy as CHD in diabetic patients may be asymptomatic or include hyperglycemia, hyperinsulinaemia and alter- present atypically (shortness of breath instead of chest ations in cell membrane electrolyte channels functions pain) than in non-diabetic patients [27]. The incidence [19]. An impaired left ventricular function occurs be- of CHD is also higher in the diabetic population more fore clinical diabetes, and affects individuals with an than in non-diabetics. In a prospective study of a dia- impaired glucose tolerance [20]. As a result, diabetics betic cohort aged >65 years and followed up for six may be more prone to HF and other cardiovascular years after first diagnosed as having diabetes, approxi- events, independently of traditional cardiovascular risk mately 40% were shown to have HF compared to 20% factors. for controls. Similarly, the rate of MI among the dia- To summarize: betic cohort was almost twice that for controls [28]. • The effects of ageing include: • Increased arterial wall thickness and stiffness • Predisposition to systolic hypertension and wide 5.4 Cardiovascular risks pulse pressure • Loss of elastic recoil of aorta and impaired coro- The National Cholesterol Education Program (NCEP) nary filling recommends that diabetic patients do not need specific • Predisposition to orthostatic hypotension CHD risk assessment, but instead that they be managed • Hypertrophy and diminished compliance of the as if they had CHD [29]. Consequently, by definition: ventricles • Relative risk is a measure of the likelihood of occur- • The effects of diabetes include: rence of the target event (death or disease) in those • Structural arterial wall changes and endothelial exposed in comparison to those not exposed to the dysfunction agent of interest. • Predisposition to a procoagulant state • Absolute risk is the likelihood or probability of oc- • Diastolic dysfunction and diabetic cardiomyopathy currence of an event (death or disease) after exposure • Increased ventricular mass to an agent (risk factor) of interest. 54 CH 05 CORONARY HEART DISEASE • Intervention is the application of a modifying agent reduction in these patients. In a small randomized con- (e.g., drug or procedure) of the baseline risk of trolled trial of patients with type 2 diabetes, when a individual or population. structured multifactorial intervention management (including behaviour modification and tight targets for • Relative risk reduction is the percentage reduction in blood glucose, blood pressure and lipids) in a spe- occurrence of an event in the population exposed to cialist setting was compared with a conventionally intervention, compared to that in the population not managed group receiving the usual care in a primary exposed to intervention (control). care setting, the risks of CVD were reduced by 0.47 • Absolute risk reduction is the percentage reduction (95% CI: 0.24 to 0.73) in the multifactorial interven- in occurrence of the event in the population exposed tion group after eight years of follow up [33]. This to intervention from their baseline risk. comprehensive approach is currently suboptimal. In a recent study to assess whether elderly patients with • Number needed to treat (NNT) is the number of type 2 diabetes (n = 48 505; aged >66 years) could patients needed to treat to prevent the occurrence use a comprehensive cardioprotective regimen (CCR) of one event over a certain period of time. of anti-hypertensive, lipid-lowering and anti-platelet Although the relative risk reduction might be similar drugs in the year following oral anti-diabetic drug ini- in both diabetics and non-diabetics, the absolute risk tiation, only 9912 (20.4%) used a CCR during the year reduction is likely to be more significantly beneficial following the first anti-diabetic medication [34]. in diabetics due to their higher baseline risk, while the number needed to treat would be expected to be lower. 5.4.1 Behaviour modification As a result, the aggressive control of all risk factors Behaviour modifications include changes in diet, is especially important in diabetics, and includes exercise, weight reduction and smoking cessation. both lifestyle modification and pharmacological Smoking cessation may be the single most effective intervention. This aggressive treatment is appropriate means of reducing mortality in high-risk populations for elderly diabetics with a life expectancy of at least [35], as smoking is known to induce vasoconstrictive 10 years, but for those with a limited life expectancy, and toxic effects on the endothelium. After only one or multiple comorbidities, the objectives should be year of smoking cessation, the excess risk associated more conservative [30]. Between 50% and 75% of with current smoking was reduced by over 50%. all deaths among patients with diabetes mellitus are However, many years of abstinence are required to cardiovascular-related, and type 2 diabetes increases reduce the risk of an ex-smoker to the level of a the risk of death from CHD by between twofold and non-smoker. A diet that is high in fibre and potassium, fourfold [31]. In The Cardiovascular Health Study, and lower in saturated fats and refined carbohydrates diabetes and increasing age were each seen to be and salt, will improve the lipid profile and significantly independent predictors of CHD fatality [Odds ratio lower the blood pressure [36]. The achievement of an (OR) 1.66, 95% CI: 1.10 to 2.31 and 1.21 per 5 years; ideal body weight through diet changes and exercise 95% CI: 1.07 to 1.37, respectively] among 5888 will reduce the overall cardiovascular risk and have a adults aged >65 years and followed up for a median favourable effect on the metabolic profile of lipids, of 8.2 years. Of the traditional CHD risk factors glycaemia and blood pressure. (hypertension, cholesterol, obesity), only diabetes was found to be an independent predictor for case fatality in this population [32]. 5.4.2 Metabolic syndrome The major cardiovascular risks include traditional Cardiovascular risk factors rarely occur in isolation risk factors of smoking, hypertension, dyslipidaemia but rather tend to cluster in what is known as in addition to hyperglycemia, visceral obesity, in- the metabolic syndrome. This is characterized by sulin resistance and hyperinsulinaemia. The identifi- a group of risk factors including visceral obesity, cation of these risk factors is vitally important in the dyslipidaemia [low high-density lipoprotein (HDL) initial evaluation of diabetic patients. Hyperglycemia cholesterol, high triglycerides], hypertension, impaired should not be treated in isolation, but the holistic glucose/insulin homeostasis (insulin resistance, view of the collective cardiovascular risk should con- hyperinsulinaemia, glucose intolerance), increased stitute a comprehensive plan of intervention and risk cardiovascular oxidative stress, impaired endothelial 5.4 CARDIOVASCULAR RISKS 55

100% 86.0%

75% 71.3%

50% 33.1% 25.8% 25%

0% Metabolic Syndrome Prevalence NFG IGT IFG DM % of Population = 56.9% 13.7% 12.3% 17.1%

Figure 5.2 Age-adjusted prevalence of metabolic syndrome in the U.S. population over 50 years of age, categorized by glucose intolerance (NFG = normal fasting glucose; IGT = impaired glucose tolerance without impaired fasting glucose; IFG = impaired fasting glucose with or without impaired glucose tolerance; DM = diabetes mellitus). Reproduced with permission from Ref. [47]. function and abnormal coagulation and fibrinolytic normal fasting glucose rising to 86% in those with profiles. Depending on various combinations of these diabetes [47] (Figure 5.2). risk factors, five definitions of metabolic syndrome The metabolic syndrome is increasingly recognized exist [37–41]. The World Health Organization as a risk factor for CVD [48], and was associ- definition includes microalbuminuria (30–300 mg per ated with ischemic electrocardiographic changes in 24 h) as a final component [37]. Microalbuminuria is 2274 elderly subjects enrolled in the Rancho Bernardo a significant marker of CVD, and is highly associated cross-sectional study [49]. The syndrome may also with hypertension and diabetes [42]. The prevalence have adverse effects on the structural and functional of metabolic syndrome increases with age; in a properties of the arteries, such as increasing arterial cohort of 2175 older people (>65 years) from the wall stiffness and thickness, through a synergistic effect Cardiovascular Health Study, the prevalence was of the clustering of its components [50]. In a prospec- 21–28% (depending on the definition used) [43]. In tive study of 888 subjects aged 40–79 years, metabolic the Three City Study, which included 5585 French, syndrome conferred a significantly increased risk for non-institutionalized, non-diabetic elderly subjects developing new carotid plaques [hazard ratio (HR) aged 65–85 years, the prevalence of metabolic 1.5], new carotid stenosis (HR 2.5) and new coronary syndrome was 12.1% [44]. In a Norwegian study, events (HR 2.3) [51]. In the Three City Study, sub- the prevalence of metabolic syndrome was increased jects with metabolic syndrome had higher frequency of from 11.0% in the 20–29 year-old group to 47.2% carotid plaques (OR 1.30, 95% CI: 1.09 to 1.55) and in the 80–89 year-old group in men, and from 9.2% higher intima-media thickness of the common carotid to 64.4% in the corresponding age groups in women artery (OR 1.81, 95% CI: 1.37 to 2.41) [44]. However, [45]. However, the magnitude of metabolic syndrome in a recent prospective study of 1025 elderly subjects in elderly diabetic individuals is higher. For example, aged between 65 and 74 years, metabolic syndrome in a population-based study of 5632 Caucasians (aged was shown to be a marker of CVD, but not above and 65–84 years), prevalences of 64.9% and 87.1% were beyond the risk associated with its individual compo- found in diabetic men and women, respectively, nents [52]. In a more recent analysis of the outcome of compared to 25.9% and 55.2% in non-diabetics [46]. two prospective studies in an elderly population aged The prevalence of metabolic syndrome also increases >60 years, metabolic syndrome and its components with increasing glucose intolerance. In the Third were associated with type 2 diabetes but showed only National Health and Nutrition Examination survey a modest association with vascular risk in elderly pop- (NHANES III) of the United States (US) population ulations. Therefore, metabolic syndrome in the elderly aged ≥50 years, there was a stepwise increase in the may not enhance risk prediction, and the criteria of prevalence of metabolic syndrome with worsening metabolic syndrome may not offer more than the sum glucose tolerance, from almost 26% in those with of its components [53]. 56 CH 05 CORONARY HEART DISEASE 5.4.3 Dyslipidaemia Although statins are known to reduce the proportional risk to equal effect in older and younger patients, Cholesterol levels tend to fall normally with increas- only limited data are available for elderly patients with ing age, and low cholesterol levels are often associated type 2 diabetes. In the CTTC meta-analysis, which in- with increased non-cardiovascular mortality, particu- cluded 18 686 patients with diabetes out of a total of larly cancer [54]. Low cholesterol levels appears to be 90 056 participants, there was a 21% reduction (95% associated with a lower body weight, disability, infec- CI: 19–23) in major vascular events per 1 mmol l−1 re- tions and other markers of general ill health, such as duction in LDL-cholesterol, and no difference in treat- low serum albumin and iron [55]. This may be related ment effect between patients with and without diabetes to reduced hepatic cholesterol synthesis, poor appetite [61]. The heart protection study included a total of 20 and low food intake in frail, older people [56]. Low 536 patients between the ages of 40–80 years, among cholesterol and low albumin levels could also be taken which the older patients (>70 years) numbered 5806 as markers of poor general health. In other words, a (28%), and total diabetics 5963 (29%). The reduction low plasma cholesterol in an elderly subject does not in cardiovascular events following simvastatin therapy necessarily cause premature death, but instead may re- (40 mg daily) was 25% after five years of follow up flect the presence of subclinical disease. In contrast, in all subgroups, irrespective of the cholesterol level at a high cholesterol level is associated with increased the start of treatment. Although the relative risk reduc- cardiovascular events and mortality in all age groups. tion was similar in all subgroups, the absolute benefit Indeed, there is a positive relationship between serum was seen to depend on the individual’s baseline risk, cholesterol level and cardiovascular risk. Statins are which is higher in diabetics [62, 63]. In a meta-analysis effective in lowering cholesterol, and reducing the risk of 12 studies, conducted to evaluate the clinical benefit of cardiovascular events, their efficacy having been of lipid-lowering in patients with and without diabetes demonstrated in numerous clinical trials with high-risk mellitus, the risk reduction for major coronary events individuals having both high and normal cholesterol was 21% (95% CI: 11–30%, P < 0.0001) in diabetic levels. The benefit of cholesterol-lowering therapy con- patients and 23% (95% CI: 12–33%; P = 0.0003) in tinues to lower levels than the normal range. In fact, non-diabetic patients for primary prevention. In sec- there is no evidence of any threshold below which ondary prevention, the corresponding risk reductions lower cholesterol levels are not associated with a lower were 21% (95% CI: 10–31%; P = 0.0005) and 23% risk, which suggests that cholesterol reduction would (95% CI: 19–26%, P ≤ 0.00001). However, the abso- be useful for all those individuals at high cardio- lute risk difference was threefold higher in the sec- vascular risk, regardless of their baseline cholesterol ondary prevention. When results were adjusted for level. baseline risk, however, the benefit was greater in di- While the evidence for cholesterol lowering is clear abetic patients, and blood lipids were reduced to a for individuals aged up to 80 years, for older pa- similar degree in both groups [64]. The post hoc anal- tients there is also some evidence of benefit from ysis of the Collaborative Atorvastatin Diabetes Study observational studies [57–59]. Whereas no mortality (CARDS) compared the efficacy and safety of ator- benefit was identified in patients aged >80 years and vastatin among 1129 patients aged 65–75 years at treated with a statin, those aged 65–79 years showed randomization with 1709 younger patients without ele- a significant (11%) reduction in mortality. There was vated LDL-cholesterol levels. Treatment with atorvas- evidence, however, of a greater trend towards bene- tatin at 10 mg per day resulted in a 38% reduction fit in those aged 80–85 years versus those aged >85 in relative risk (95% CI: –58 to –8, P = 0.017) of years [59]. The positive association between total and the first major cardiovascular event in older patients, LDL-cholesterol and cardiovascular risk becomes at- and a 37% reduction (95% CI: –57 to –7, P = 0.019) tenuated with advancing age (and more so in men than in younger patients. The corresponding absolute risk in women [60]). However, in the Cholesterol Treatment reductions were 3.9 and 2.7%, respectively (differ- Trialists Collaborators (CTTC) systematic prospective ence 1.2%, 95% CI: –2.8 to 5.3; P = 0.546), while meta-analysis, which reported data from 14 random- the numbers needed to treat (NNT) for four years to ized trials, those patients aged >65 years (n = 6446) avoid one event were 21 and 33, respectively. The had 19% reduction in the risk of major cardiovascular higher absolute risk reduction and lower NNT in the events, a benefit similar to the 22% reduction in risk elderly reflect their higher baseline risk. All-cause mor- experienced by those aged <65 years (n = 7902) [61]. tality was reduced non-significantly, by 22% (95% CI: 5.4 CARDIOVASCULAR RISKS 57

Atv Pbo RRR n (%) n (%) (%) Primary composite end point Total population 83 (5.8) 127 (9.0) 37* Age ≥65 41 (7.2) 62 (11.1) 38* Age <65 42 (7.9) 65 (7.6) 37* Test of heterogeneity: P = 0.963

Acute coronary event Total population 51 (3.6) 77 (5.5) 36* Age ≥65 26 (4.5) 37 (6.6) 34 Age <65 25 (2.9) 40 (4.7) 39 Test of heterogeneity: P = 0.807

Coronary revascularization Total population 24 (1.7) 34 (2.4) 31 Age ≥65 6 (1.0) 13 (2.3) 56 Age <65 18 (2.1) 21 (2.5) 15 Test of heterogeneity: P = 0.253

Fatal and nonfatal stroke Total population 21 (1.5) 39 (2.8) 48* Age ≥65 13 (2.3) 24 (4.3) 49 Age <65 8 (0.9) 15 (1.8) 48 Test of heterogeneity: P = 0.966

0.0 0.50 1.0 1.5 2.0 Relative risk reduction Atorvastatin better Placebo better *p <0.05 for atorvastatin vs. placebo.

Figure 5.3 Composite primary end point and components. The total number of acute coronary events, coronary revascularizations and strokes (separately) do not equal the total number of primary events shown above, because only the ﬁrst of these events is included in the primary end point. Atv = atorvastatin; Pbo = placebo; RRR = relative risk reduction. Reproduced with permission from Ref. [65].

–49 to 18, P = 0.245) and 37% (95% CI: –64 to 9, the risk of serious muscle adverse effects may be P = 0.98), respectively. The reduction in total choles- slightly higher. Higher doses of statins should be used terol, LDL-cholesterol and triglyceride was similar in with caution in frail elderly patients, who may be both age groups, as was the overall safety profile of more susceptible to drug-related myopathy and other atorvastatin. The authors concluded that the absolute side effects, as statin toxicity has been shown to be and relative benefits of statin therapy in older patients dose-related. with type 2 diabetes were substantial, and that all older patients warranted treatment, unless specifically con- 5.4.4 Hypertension traindicated [65] (Figure 5.3). It appears from the above that statins should be The prevalence of hypertension in diabetics is up prescribed for all older people with diabetes who have to threefold that in non-diabetics [67]. Hypertension a reasonable life expectancy. Chronological age, in and affects up to 60% of patients with type 2 diabetes of itself, should not exclude patients from receiving [68], with increasing age, obesity and the onset of renal therapy, while the functional or biological age of the disease being the contributory factors to increasing its patient, and the impact of long-term drug therapy prevalence among these patients. The development of on safety and quality of life, should be considered. type 2 diabetes is also about twice as likely in persons Given the larger reduction in event rates among older with hypertension than in normotensive people, which patients, treatment would also be expected to be more suggests a frequent coexistence of these two common cost-effective in older than in younger patients [66]. chronic diseases [69]. Hypertension markedly increases Moderate-dose statins appear to be well tolerated in the risk of CVD in patients with type 2 diabetes elderly persons participating in clinical trials, although compared to those without diabetes [67]. 58 CH 05 CORONARY HEART DISEASE Hypertension should be managed aggressively in factor, and were without HF or a low ejection fraction. diabetic patients. A tight blood pressure control be- Patients were randomized to ramipril (10 mg/day) or low the recommended levels of 140/90 mmHg [70] placebo for four-and-a-half-years of follow up. Those markedly reduced not only CVD but also the devel- assigned to ramipril had fewer major vascular events opment of end-stage renal disease in persons with compared to those receiving to placebo (18.6 versus type 2 diabetes mellitus. This risk reduction was even 24.0%, HR = 0.75, P = 0.0006), cardiovascular deaths more impressive than the tight blood glucose control (9.3 versus. 13.0%, HR = 0.71, P = 0.003), MI (12.0 in the United Kingdom Prospective Diabetes Study versus 15.6%, HR = 0.75, P = 0.006) and strokes (5.4 (UKPDS), where the benefits of a diastolic blood pres- versus 7.7%, HR = 0.69, P = 0.013). These were of sure reduction to 82 mmHg in the tight control group similar magnitude to the proportional reductions in compared to 87 mmHg in the usual care group, dramat- risk observed in patients aged <70 years. Ramipril ically outweighed those of intensive glucose control was generally safe, and equally well tolerated in pa- of mean hemoglobin A1c (HbA1c) reductionto7.0% tients aged ≥70 years and <70 years. Importantly, versus 7.9% in both groups, respectively [71]. The due to the high risk of cardiovascular events in el- reduction of diastolic blood pressure to <80 mmHg derly patients, the absolute risk reductions attained with reduced CVD events by 51% in comparison to a dias- ramipril were higher in patients aged ≥70 years than tolic blood pressure of 90 mmHg in the diabetic sub- in those aged <70 years, for most endpoints. As an group of the Hypertension Optimal Treatment (HOT) example, the absolute risk reduction for the primary study [72]. In contrast, those HOT study participants endpoint was 5.4% in patients aged ≥70 years and 3% without diabetes received no benefit from this further for those aged <70 years; consequently, for elderly diastolic blood pressure reduction. The reduction of patients the NNT to prevent one major cardiovascu- systolic pressure, from 175 to 153 mmHg, led to a sig- lar event over four-and-a-half years was 18, compared nificant reduction in CVD-related events in the Systolic to 33 for younger patients [76]. These cardiovascular Hypertension in Europe (Syst-Eur) trial [73]. Here, di- and renoprotective beneficial effects of ACE inhibitors abetic patients showed more benefit from an aggressive in patients with type 2 diabetes were independent of blood pressure lowering than did non-diabetics. In that any blood pressure lowering, and may occur whether trial, despite the systolic pressure being reduced by a albuminuria is present, or not [77, 78]. comparable amount in each group (22.0 ± 16 mmHg Beta-blockers are associated with an increased risk in non-diabetics versus 22.1 ± 14 mmHg in diabet- for new-onset diabetes mellitus, with no benefit for ics), the risk reduction in mortality from CVD was the end point of death or MI, and with a 15% in- 13% in non-diabetic patients compared to 76% for di- creased risk for stroke compared to other agents. A abetics. The same effect was apparent in the Systolic meta-analysis of 12 studies evaluating 94 492 patients Hypertension in the Elderly Program (SHEP) study, taking beta-blockers as first-line therapy for hyperten- where elderly persons with type 2 diabetes derived sion with data on new-onset diabetes and a follow more benefit from an aggressive systolic blood pressure up for more than one year showed beta-blocker ther- lowering in reduction of CVD than did those without apy to result in a 22% increased risk for new-onset diabetes [74]. Based on the above findings, it appears diabetes (relative risk 1.22, 95% CI: 1.12 to 1.33) that the benefit per mmHg blood pressure reduction is compared to non-diuretic antihypertensive agents. A greater in diabetic patients than in those who are hyper- higher baseline fasting glucose level was a significant tensive but have no concomitant diabetes mellitus, and predictor of new-onset diabetes (OR 1.01, 95% CI: this confirms the need for an aggressive reduction of 1.00 to 1.02, p = 0.004). The risk for diabetes was arterial pressure in diabetic patients. In order to achieve greater with atenolol in the elderly and in studies in these goals, most older diabetic people will require a which beta-blockers were less efficacious antihyper- combination of at least two or three antihypertensive tensive agents, and increased exponentially with an in- medications [75]. creased duration of beta-blocker therapy. On the other Angiotensin-converting enzyme (ACE) inhibitors hand, calcium channel blockers (CCBs) and ACE in- may have a role in reducing cardiovascular risk in hibitors or angiotensin II receptor blockers (ARBs) re- older diabetics. The elderly Heart Outcomes Prevention sulted in 21% and 23% reductions, respectively, in the Evaluation (HOPE) trial included 2755 older patients risk for new-onset diabetes compared to beta-blockers who were aged ≥70 years with vascular disease or di- [79]. Also, in comparison with other agents, the an- abetes, had at least one additional cardiovascular risk tihypertensive efficacy of beta-blockers was inferior. 5.4 CARDIOVASCULAR RISKS 59 In this analysis, however, diuretics resulted in an death from stroke or from any cause, although only increased risk for new-onset diabetes compared to approximately 7% of the subjects were diabetics [89]. beta-blockers, although their blood pressure-lowering In summary, risk reduction via hypertension control efficacy was superior. In the UKPDS, although atenolol in patients with diabetes is substantially greater than efficacy was similar to that of the ACE inhibitor, cap- that in non-diabetic persons, who have similar blood topril, those patients receiving atenolol gained more pressure levels [72]. Blood pressure in diabetics should weight and required a more frequent addition of new be treated aggressively, and indeed the control of this glucose-lowering agents than those taking captopril parameter may be the most important factor in pre- [71]. A recent meta-analysis showed that the associ- venting adverse outcomes. Most patients will need to ation of antihypertensive drug class on incident dia- be administered more than one antihypertensive agent, betes was lowest for ARBs, ACE inhibitors followed and in this respect thiazide diuretics, ARBs, ACE in- by CCBs, beta-blockers and diuretics, in rank order hibitors and CCBs seem to be reasonable first-choice [80]. Beta-blockers worsen glycemic control by in- agents, although higher doses of diuretics may worsen hibiting pancreatic insulin secretion, increasing both blood glucose and lipid levels [90]. The addition of insulin resistance and glycogenolysis [81]. However, other agents may be necessary in order to achieve the newer, non-cardioselective beta-blockers with va- blood pressure targets. sodilating properties (e.g., carvedilol) have minimal effects on glycemic control [82]. 5.4.5 Hyperglycemia The Antihypertensive and Lipid-Lowering treatment While diabetes has long been recognized as a risk to prevent Heart Attack Trial (ALLHAT) compared factor for CHD, it is not clear whether hyperglycemia ACE inhibitors, CCBs and thiazide diuretics. In a itself is a risk for CVD. Hyperglycemia leads subgroup analysis of 12 063 patients with type 2 to an increase in oxidative stress, an enhanced diabetes, no significant differences were seen between leukocyte–endothelial interaction and the glycosyla- different groups in the primary outcomes of non-fatal tion of proteins in the body, including lipoproteins, MI plus CHD death or all-cause mortality. However, apolipoproteins and clotting factors [91]. This may the risk for HF was lowest in the diuretic group [83]. have a role in increasing cardiovascular risk. In com- The ARBs reduce not only renal end points but also bination with other risk factors, hyperglycemia may cardiovascular events [84–86]. In the Losartan Inter- cause an accelerated progression of atherosclerosis in vention for Endpoint Reduction (LIFE) study, patients people with diabetes, although few data are available with hypertension and signs of left ventricular hyper- to assess whether lowering the blood glucose level trophy on electrocardiography were assigned randomly will predict CVD outcome. Prospective studies have to an ARB (losartan) or a beta-blocker (atenolol). In a demonstrated that glucose is a continuous CHD risk subgroup analysis of 1195 patients with diabetes, the factor in diabetics [92, 93], the results having sug- losartan group had a substantially lower risk for cardio- gested that the risk of CHD rises by 10–30% for each vascular end-points and total mortality [87]. A recent 1% increase in HbA1c level. In the UKPDS, intensive meta-analysis to compare the benefits and adverse ef- treatment with metformin in a subgroup of patients fects of ACE inhibitors versus ARBs for treating essen- with type 2 diabetes resulted in a decreased risk for tial hypertension showed that both agents had similar MI and stroke compared to the conventional treatment effects on blood pressure control, but ACE inhibitors group, while intensive treatment with sulfonylurea or caused higher rates of cough than did ARBs. The data insulin improved the long-term outcomes of diabetes regarding other outcomes were limited [88]. (microvascular and macrovascular complications); The above evidence applies to older people (up to however, the improvement in cardiovascular risk did the age of 80 years), as clinical trials have either ex- not reach statistical significance [94]. In the same cluded this group of the population, or included only study, the data also indicated that with each 1% rise very few. The recently published Hypertension in the in HbA1c the incidence of MI rose by about 14%. Very Elderly Trial (HYVET) has provided evidence Moreover, the relationship between glycaemia and that antihypertensive treatment with the diuretic inda- cardiovascular risk was seen to start within the normal pamide (sustained release), with or without the ACE blood sugar range, with a linear relationship, and inhibitor perindopril, in persons aged ≥80 years, is showed no indication of any threshold effect. This beneficial and associated with reduced risks of HF, suggested that CHD started before the onset of clinical 60 CH 05 CORONARY HEART DISEASE diabetes [95, 96]. A meta-analysis of three studies their expression and intensities may vary between races involving persons with type 1 diabetes (n = 1688), and ethnic groups. In southern Asia, for example, in and 10 studies involving persons with type 2 diabetes recent years there has been a progressive increase in (n = 7435), showed a positive effect of blood glucose the prevalence of diabetes mellitus and CHD due to lowering on the risk of cardiovascular events, and the lifestyle changes, modernization and increasing urban- effect seemed to be greater in type 2 diabetics than in ization and industrialization, all of which have led to type 1. For each 1% increase in HbA1c the relative physical inactivity and obesity [103]. Yet, the preva- risk of CVD was 1.18 (95% CI: 1.10 to 1.26) for type lence of diabetes has also increased in southern Asians 2 diabetes and 1.15 (95% CI: 0.92 to 1.43) for type who have migrated to western countries, with south- 1 diabetes [97]. The relationship of postprandial gly- ern Asians living in the UK having a fourfold higher caemia, fasting blood glucose and cardiovascular risk prevalence of diabetes than their British counterparts in individuals with diabetes was less clear, although [104]. Southern Asians are also genetically prone to most reports have indicated a greater pathogenic develop both diabetes and CHD, and this tends to potential of postprandial hyperglycemia rather than occur about a decade earlier than in western coun- fasting hyperglycemia [98]. A meta-analysis of ran- tries [105]. In addition to the ‘traditional’ risk fac- domized trials targeting postprandial hyperglycemia tors, southern Asians may have a higher prevalence showed cardiovascular benefit in type 2 diabetes. of emerging and new atherogenic risk factors, such as The treatment of postprandial hyperglycemia was lipoprotein abnormalities and hyperhomocysteinaemia, associated with a reduction in the development of that make them more prone to develop CVD [106]. any CVD by 35% [99]. These observations support This higher risk should be considered when south- the hypothesis that the lowering of blood glucose to ern Asians living in the ‘developed’ world are eval- levels within the normal range may either prevent or uated. postpone the development of CHD. However, as there Black race is associated with increased mortality are at present no data available specifically for older from ischemic heart disease compared to whites, and diabetics, evidence must be extrapolated to the older risk factors such as diabetes and hypertension tend age groups. to be less well controlled in blacks than in whites [107, 108]. Although this might lead to an increased cardiovascular risk in black diabetics, the evidence is 5.4.6 Hypercoagulability conflicting. While one American study [109] reported a A procoagulant state has been demonstrated in indi- greater CVD prevalence among black geriatric patients viduals with diabetes [100]. Both, platelet aggregation with diabetes, others have not found this to be the case and adhesion are increased in diabetics [101]. Diabetes [110–112]. Black diabetics of Afro-Caribbean descent increases intrinsic platelet activation and decreasing who are resident in the UK have also been found to the endogenous inhibitors of platelet activity [102]. be at a lower risk of CVD than white Europeans with These changes are likely due to a chronic inflammatory diabetes [113]. In a recent study, there was no signif- state induced by diabetes. Consequently, older diabet- icant difference in the incidence of CVD among dia- ics should be prescribed antiplatelet treatment as part betic elderly black (23.9%) and white (29.2%) Amer- of a multifactorial treatment to reduce their cardiovas- icans. However, the risk of CVD was lower among cular risk, regardless of whether they have a history of southern black men (HR 0.87, 95% CI: 0.82–0.92) CHD, or not. and women (HR 0.95, 95% CI: 0.91–0.99) than their southern white counterparts. In the three other US re- 5.4.7 Race and ethnicity gions combined (northeast, midwest and west), black men had a similar risk for CVD (HR 1.01, 95% CI: Most present-day knowledge regarding cardiovascular 0.95–1.07), while black women had a greater risk (HR risk is derived from the Framingham population, the 1.10, 95% CI: 1.05–1.16) than non-southern white men majority of which was white people of European ori- and women, respectively. The greater risk for CVD of gin. It is uncertain whether baseline absolute risk is black women with diabetes not residing in the south similar to that in other populations, but race and eth- (versus white women with diabetes) argues against any nicity will have a significant influence on the risk of genetic explanation, and suggests that regional differ- CVD. Although, the most important risk factors asso- ences in lifestyle, environment or quality of care may ciated with diabetes are largely similar in all countries, play some role [114]. 5.5 MYOCARDIAL INFARCTION 61 To summarize: warrant early investigation. However, exercise tolerance test or stress echocardiography are not routinely • Absolute risk reduction is higher and more cost recommended. A 12-lead resting electrocardiogram has effective in diabetics than in non-diabetics. low sensitivity and specificity, but is a useful baseline • Metabolic syndrome in the elderly may not enhance [119]. The current recommendation is that patients with the prediction of risks above its components. diabetes and two additional risk factors should undergo stress myocardial perfusion scans, although most el- • Statin therapy is equally beneficial in older and derly diabetics will have multiple risk factors for CHD younger diabetics. and the burden of such screening would not be practi- • Blood pressure control has a greater impression on cal [120]. Although cardiovascular risk factors should risk reduction than does blood glucose control. be treated in older diabetics on the basis of the diabetes alone, screening may be indicated in higher risk • The diabetic elderly benefit more than non-diabetics patients in whom revascularization therapy is thought from tight blood pressure control. to be indicated. • The benefits of blood pressure reduction are similar The increased risk for CHD by diabetes appears to for patients of all age groups. be greater in women. In a meta-analysis to estimate the relative risk for fatal CHD associated with diabetes • All classes of antihypertensive medication are effec- in 37 studies, which included a total of 447 064 pa- tive in both elderly and younger people. tients (some studies included people aged >65 years), • Southern Asians demonstrate a higher risk, at a the rate of fatal CHD was higher in diabetic than in younger age, than do whites. non-diabetic patients (5.4% versus 1.6%). The overall summary relative risk for fatal CHD in patients with • Hyperglycemia should not be treated in isolation, but diabetes compared with no diabetes was significantly as part of multifactorial intervention. greater among women than it was among men (3.50, 95% CI: 2.70–4.53 versus 2.06, 95% CI: 1.81–2.34). 5.5 Myocardial infarction Following exclusion of the eight studies that had adjusted only for age, the difference in risk between the The prevalence of CHD may be as high as 55% among genders was substantially reduced, but still highly sig- adults with diabetes compared to only about 4% among nificant. The pooled ratio of the relative risks (female: those without diabetes [115]. The risk of MI in pa- male) from the 29 studies with multiple adjusted es- tients with diabetes mellitus without a history of MI timates was 1.46 (95% CI: 1.14–1.88). The gender is as high as that in patients without diabetes mellitus difference in CHD risk might be a consequence of dia- who have had an MI [2]. Following the development betes inducing a more adverse cardiovascular risk pro- of CHD in diabetics, the prognosis is worse and mor- file in women, combined with a reduced likelihood of tality after the first MI higher than in non-diabetics women receiving recommended levels of cardioprotec- [116]. Age is the most powerful predictor of CHD tive treatment [121]. The Hoorn study also emphasized risk; by the age of 65 years, most men have a 20% the increased hazard in women with diabetes in a pop- risk of a CHD event occurring within the next 10 ulation of patients aged 62 ± 7 years [122]. Diabetes years, with diabetes conferring a risk equivalent to age- also increases CHD mortality for older people; in a ing 15 years. In men, the transition from intermediate cohort study of older patients (average age 72 years), to high-risk status occurs at an average age of 47.9 diabetes conferred a CHD mortality risk that was 2.5- years, whereas in women the transition is about seven to 2.75-fold greater than for patients without diabetes years later, at the age of 54.3 years [117]. Screening [123]. The duration of diabetes represents a risk for for CHD is based on symptoms of angina, with a low the development of CHD. For example, in the Fram- threshold for investigation. Asymptomatic MI or my- ingham study, the risk of CHD events and CHD death ocardial ischemia is more common in diabetic patients, were shown to be 1.38 (95% CI: 0.99–1.92) and 1.86 especially the elderly [118]. Patients with diabetes may (95% CI: 1.17–2.93) times greater for each 10 years also present with atypical symptoms of myocardial is- increase in the duration of diabetes, respectively [124]. chemia, such as an unexplained shortness of breath However, there is an increased risk for CHD even be- on exertion, which may be an angina equivalent and fore the clinical diagnosis of diabetes [125]. 62 CH 05 CORONARY HEART DISEASE Beta-blockers are associated with a lower one-year also tended to have more severe diffuse three-vessel mortality rate for elderly diabetic patients, similar to or left main stem disease than those without diabetes. that in non-diabetics. In a retrospective cohort study Typically, such patients showed about 30% mortality of 45 308 patients admitted to hospital with acute before receiving thrombolytic therapy, while among MI, and after adjusting for potential confounders, patients receiving thrombolysis the mortality was about beta-blockers were associated with a lower one-year twice that of non-diabetics [131]. mortality for insulin-treated diabetics (HR = 0.87, Details of the effect of glucose, insulin and potas- 95% CI: 0.72–1.07), non-insulin-treated diabetics sium (GIK) therapy on the mortality of acute MI are (HR = 0.77, 95% CI: 0.67–0.88) and non-diabetics conflicting. The Diabetes Insulin–Glucose in Acute (HR = 0.87, 95% CI: 0.80–0.94). Although some con- Myocardial Infarction (DIGAMI) study, which was re- cern was expressed regarding the use of beta-blockers ported in 1995, concluded that in diabetic patients an in diabetic patients, such therapy was not significantly insulin–glucose infusion followed by a multidose in- associated with any increase in the six-month sulin regimen for 3 months after discharge improved readmission rates for diabetic complications among long-term survival after an acute MI. At one year, diabetics [126]. mortality in the infusion group was 18.6% compared The outcome after MI is poor in older diabetics. For to 26.1% in the control group (P < 0.0273), with a example, the outcome data of 1698 elderly (aged ≥65 relative reduction in mortality of 30% [132]. The pos- years) diabetics after one year of hospitalization for itive results of DIGAMI continued for a mean of 3.4 MI showed a high risk of HF, recurrent MI and mor- years of follow up, with a 33% mortality occurring in tality. Comorbid conditions related to diabetes mellitus the insulin–glucose infusion group compared to 44% (previous HF, MI, chronic renal impairment, peripheral in the control group (P = 0.011) [133]. However, the vascular disease, stroke) at the time of the index MI DIGAMI II study, the details of which were reported were important contributors to poorer outcomes in el- in 2005, did not support the concept that both short- derly patients with diabetes mellitus [127]. A higher and long-term insulin therapy would improve survival blood glucose level (>6.1 mol l−1) was also associ- in patients with type 2 diabetes mellitus with acute ated with an increased mortality risk in elderly acute MI compared to conventional therapy [134]. The ev- MI patients, particularly those without recognized dia- idence of GIK in trials which included non-diabetic betes [128]. patients was also conflicting. For example, in 1997 a In addition to a poor outcome, older diabetics have meta-analysis of GIK therapy for acute MI showed a more comorbidities and a greater resource use than mortality of 21% in the placebo group and 16.1% in non-diabetics. In another study of older people hospi- the GIK group. The risk reduction in mortality was talized for acute MI, older diabetics constituted 33% of 28% (OR 0.72, 95% CI: 0.57–0.90, P < 0.004). How- the total admissions in one year. Comorbid conditions, ever, the studies included in the meta-analysis were including hypertension, prior acute MI, prior stroke conducted before the reperfusion therapy era [135]. Af- and/or prior revascularization, were more frequent in ter 1997, most of the clinical trials did not show any diabetic than in non-diabetic patients. Congestive heart benefit of GIK therapy on mortality [136]. failure (CHF) also occurred more frequently in diabetic To summarize: patients. The length of hospital stay (7.9 versus 7.0 • The risk of MI in diabetics is at least twice that in days, P < 0.001), in-hospital mortality rates (16% ver- non-diabetics sus 13%, P < 0.001) and rates for mortality within 30 days (21% versus 17%, P < 0.001) were each higher • Presentation may be atypical in older people in diabetic patients [129]. The long-term survival after • acute MI was also significantly decreased in diabetic The outcome of MI is worse in diabetics than in patients compared to non-diabetics. Over a 12-year pe- non-diabetics riod, the relative risk of dying was 1.56-fold higher • The medications used in MI treatment have similar among diabetic men than among non-diabetic men efficacies in diabetics and non-diabetics (95% CI: 1.43–1.68), whereas diabetic women were 1.57-fold more likely to die than non-diabetic women • The role of glucose, insulin and potassium infusion (95% CI: 1.45–1.73) [130]. Diabetics with acute MI therapy remains controversial 5.6 HEART FAILURE 63 5.6 Heart failure Few data are available about reducing the risk of developing HF in diabetics, specifically in older people. It Type 2 diabetes is associated with a greater risk of is unclear whether an improvement in glycemic control HF and, indeed, both conditions frequently coexist. would reduce the incidence of HF. While a high HbA1c The mechanisms underlying this association remain level has been shown to increase the risk of HF [139], controversial, but may include diabetes-associated co- intensive glycemic control did not significantly reduce morbidities such as hypertension and obesity, diabetes- the incidence of HF in the UKPDS [147], although in associated complications such as small and large ves- the same study a tight control of the blood pressure sel disease, and diabetic cardiomyopathy. Independent led to a reduction in the risk of developing HF (HR risk factors for the development of HF in diabetes in- 0.44, 95% CI: 0.2–0.94, P = 0.0043) [71]. Some evi- clude a higher HbA1c level [137] and an increased body dence has been produced for the use of ACE inhibitors, mass index (BMI) [137, 138]; increasing the BMI by ARBs and statins in reducing the risk of HF, with the 2.5 units increases the risk of HF by 12% [137]. In ACE inhibitor ramipril producing a 20% risk reduction = the UKPDS, for every 1% reduction in HbA1c the risk (95% CI: 4–34, P 0.019) in the MICROHOPE trial of HF fell by 16% [139]. Other factors included in- [148]. In the IDNT, the ARB irbesartan reduced the creasing age [137], CHD [137, 138], the use of insulin incidence of HF compared to placebo (HR 0.72, 95% [137], end-stage renal disease [137], microalbuminuria CI: 0.52–1.00, P = 0.048) [86], while high-dose ator- [138], retinopathy [140] and the duration of diabetes vastatin (80 mg/day) reduced the hospitalization time [137]. Diabetes also is a risk factor for the development for HF in the Treating to New Targets study [149]. of HF with a normal ejection fraction (HFNEF); this It appears, however, that no specific drug therapy, al- condition is commonly referred to as ‘diastolic HF’, ternative glycemic control or different blood pressure although most such patients also have associated sys- target is available either to prevent or to treat HF in di- tolic dysfunction. An overlap occurs between diabetes, abetic patients. Subgroup analyses of HF clinical trials hypertension and ischemic heart disease in the devel- have suggested similar treatment benefits from ACE opment of HFNEF [141]. The prevalence, incidence inhibitors and beta-blockers, regardless of the patients’ and mortality of HF in older patients with diabetes is diabetes status [150]. very high. In a population-based cohort study of 151 In diabetics, the possibility of HF should be con- 738 patients with diabetes, all of whom were aged >65 sidered in the presence of risk factors such as CHD, years, HF prevalence was 22.3% and the incidence rate hypertension, proteinuria [151] and retinopathy [140]. 12.6% (95% CI: 12.5–12.7%). Increasing age, ischemic B-type natriuretic peptide may also represent a useful heart disease, nephropathy and peripheral vascular dis- screening approach [152], although is no particular in- ease were strong predictors for the development of HF. vestigation is available for HF in diabetes which differs Over a 60-month follow-up period, however, mortality from those in non-diabetics. Neither is there any spe- among HF patients was almost 10-fold that in diabetic cific difference in the clinical presentation of HF in di- patients who remained HF free (32.7% versus 3.7%) abetics. The typical diabetic patient with HFNEF is an [142]. In a recent study, the cumulative incidence of elderly woman with a history of hypertension, in whom HF in older diabetics was 47.5% after six years of di- the HF is episodic and often precipitated by an episode agnosis of diabetes compared to 20% in a non-diabetic of atrial fibrillation, ischemia or infection [153]. control group [28]. Higher rates of HF with a greater Beta-blockers offer similar benefit for diabetics and number of complications were related to poor glycemic non-diabetics in HF, with relative risk reductions in and blood pressure control. This was consistent with mortality of 0.84 (95% CI: 0.73–0.96, P = 0.011) hav- the limited clinical trials data, which suggested that ing been shown in diabetics compared to 0.72 (95% CI: blood pressure control, treatment with ACE inhibitors 0.65–0.79, P < 0.001) in non-diabetics [154]. The ben- and glycemic control might reduce the risk of HF [141, efits in morbidity were also similar [155]. Diabetics are 143]. Ischemic heart disease and renal impairment, as less likely than non-diabetics with HF to be discharged predictors of HF, have been demonstrated previously from hospital while receiving beta-blocker therapy [144, 145]. In nursing homes, the incidence of HF in (OR 0.72, 95% CI: 0.55–0.94) [156]. Beta-blockers residents with diabetes was seen to be almost twice that are cautiously prescribed for older HF diabetics be- of those without diabetes (12.1% versus 6.2%) [146]. cause of their perceived unfavourable effects on glu- 64 CH 05 CORONARY HEART DISEASE cose metabolism and a paucity of available clinical To summarize: data. However, the results of a recent study showed that diabetes does not negatively influence the safety, • Heart failure and diabetes commonly coexist tolerability and efficacy of carvedilol in older dia- • Diabetes is associated with both reduced and normal betic patients (aged >70 years) with systolic HF. At ejection fractions in HF a one year follow up, the tolerability (93.7% versus 92.2%) and mean daily dose (24 ± 17 versus • The presence of CHD, hypertension, proteinuria and 23 ± 14 mg) of carvedilol were similar in diabetics retinopathy should lead to a suspicion of HF in and non-diabetics. Neither was any worsening of fast- diabetics ing blood glucose level, HbA and creatinine lev- 1c • els, nor of the incidence of death and hospitalization, Diabetes worsens the outcome of HF, especially in observed in diabetics treated with carvedilol. Similar older women improvements in New York Heart Association class • All HF medications have similar efficacies in both and mitral regurgitation severity were observed in di- diabetics and non-diabetics abetic and non-diabetic patients receiving carvedilol. Although beta-blockers have previously been well tolerated in older diabetic patients with HF, diabetes 5.7 Anti-diabetic medications and remains a strong prognostic factor that limits the re- CHD versibility of left ventricular systolic dysfunction and the effect of treatment on subsequent outcome [157]. An early clinical trial suggested that sulphonylureas Other medications used in HF have similar effects are cardiotoxic and may exacerbate diabetic cardiomy- in diabetics and non-diabetics. The relative risk of opathy [164], although this was not confirmed in the mortality among ACE inhibitor-treated, compared to UKPDS [152]. In a retrospective cohort study, patients placebo-treated, patients was 0.85 (95% CI: 0.78–0.92) commenced on insulin had a higher incidence of HF in non-diabetics and 0.84 (95% CI: 0.70–1.00) in dia- hospitalization than those commenced on sulphony- betics [158]. A similar pattern was shown with ARBs lureas (HR 1.56, 95% CI: 1.00–2.45, P = 0.05) [165]. therapy [159, 160]. Spironolactone has similar mortal- In another retrospective cohort study, there was no ef- ity benefits in diabetics (HR 0.70, 95% CI: 0.52–0.94, fect of sulphonylureas on mortality (HR 0.99, 95% CI: = P 0.019) and non-diabetics (HR 0.70, 95% CI: 0.91–1.08) [166]. Although insulin therapy has been 0.60–0.82, P < 0.001) with severe HF [161]. No data shown to predict the development of HF and mortal- are presently available to suggest that other medica- ity in diabetes [167, 168], this was not shown in the tions used in HF treatment (e.g., diuretics, digoxin, UKPDS. However, insulin use is likely to be started nitrates, hydralazine) will have any different effects at a late stage in type 2 diabetes, when macrovascular in diabetes. Diabetes worsens the outcome of both disease could have been already established. systolic and diastolic HF in older people, especially Cardiovascular outcome seems to be better among in women [162]. In the Digitalis Investigation Group patients receiving metformin therapy. In the UKPDS, trial, diabetes-associated increases in hospitalization in a subgroup of overweight subjects with type 2 and mortality in chronic HF were more pronounced in women, and these gender-related differences in out- diabetes mellitus, metformin treatment caused a de- come were primarily observed in elderly patients (aged crease in all-cause mortality, notably that due to MI ≥65 years). The absolute increase in all-cause hospi- [169]. In patients with newly diagnosed HF, metformin talizations due to HF in women was 74% and 61%, monotherapy was associated with a reduced one-year respectively, for patients with and without diabetes mortality when compared to sulfonylurea treatment (p < 0.001), while the absolute increase in mortality (HR 0.66, 95% CI: 0.44–0.97) [170]. The one-year in women was 39% and 25%, respectively, for pa- mortality was also lower in patients receiving a met- tients with and without diabetes (p < 0.001). These formin/sulfonylurea combination therapy than in those findings suggest that diabetes has a significant nega- with sulfonylurea monotherapy (HR 0.54, 95% CI: tive impact on the natural history of HF, and an early 0.42–0.70). Metformin use was also associated with diagnosis and a more tight control of diabetes is neces- a lower one-year mortality compared to insulin or sary to improve outcomes in HF, especially in elderly sulphonylureas treatment in patients admitted to hos- women [163]. pital with HF (24.7 versus 36%, P < 0.0001) [166]. It 5.7 ANTI-DIABETIC MEDICATIONS AND CHD 65 has been suggested that metformin be used with cau- P < 0.001), acute MI (RR 1.40, 95% CI: 1.05–1.86, tion in HF patients due to a risk of lactic acidosis, P = 0.02) and death (RR 1.29, 95% CI: 1.02–1.62, although such risk does not appear to be high. Indeed, P = 0.03) The increased risks associated with TZDs in a retrospective cohort study the rate of lactic acido- were independent of baseline cardiovascular risk or di- sis was 2.3% in metformin-treated patients compared abetes duration. The increased risk of CHF, acute MI to 2.6% in those not treated (P = 0.40) [166]. A similar and mortality associated with TZDs use appeared lim- retrospective study reported no cases of lactic acidosis ited to rosiglitazone [179]. A recent meta-analysis of with metformin in diabetics with HF [170]. 42 trials comparing rosiglitazone with placebo or active The thiazolidinediones (TZDs), rosiglitazone and comparators in more than 27 000 patients with dia- pioglitazone, have been shown to improve glycemic betes suggested that treatment with rosiglitazone was control and to slow the progression of beta-cell fail- associated with an increased risk of MI and cardiovas- ure [171]. These drugs have multiple therapeutic ef- cular death [176]. Another meta-analysis to evaluate fects through their action on peroxisome proliferator the effect of pioglitazone on ischemic cardiovascular activated receptor gamma, which is expressed dif- events included a total of 19 trials enrolling 16 390 fusely in human tissues. Their beneficial effects include patients. Death, MI or stroke occurred in 4.4% of pa- reductions in insulin resistance and hyperglycemia, tients receiving pioglitazone and in 5.7% of patients re- anti-inflammatory effects and improvement of hyper- ceiving control therapy (HR 0.82, 95% CI: 0.72–0.94, tension, microalbuminuria and hepatic steatosis. These P = 0.005). The individual components of the primary TZDs effects may be independent of their effects on blood glucose. While improved glycemic control has end point were each reduced by a similar magnitude been linked to better clinical outcomes [172] and TZDs with pioglitazone treatment, with HRs ranging from have been suggested as having potential cardiovascu- 0.80 to 0.92. However, serious HF was reported in lar benefits [173], this group of drugs has recently been 2.3% of the pioglitazone-treated patients and in 1.8% scrutinised for increased cardiovascular risk. TZDs are of the control patients (HR 1.41, 95% CI: 1.14–1.76, associated with weight gain, oedema [174] and an in- P = 0.002) (Table 5.1). creased risk of CHF [173]. The frequency of oedema It appeared that pioglitazone was associated with a is approximately 5% when TZDs are used either in significantly lower risk of death, MI or stroke among a monotherapy or as combination oral therapy, and ap- diverse population of patients with diabetes. However, proximately 15% when used with insulin [175]. The serious HF was increased by pioglitazone, though with- mechanism of fluid retention is not clear, and is largely out any associated increase in mortality [173]. The rea- peripheral, but may result from changes in hemody- son why these two TZDs have different effects on car- namics, with some contribution from molecules that diovascular outcome is unclear, but it may be related to regulate cell and tissue permeability. There may also the fact that pioglitazone produces greater reductions in be a direct effect of TZDs on sodium reabsorption via serum triglycerides and increases in HDL-cholesterol the renal medullary collecting ducts [175]. The results levels [180]. Although the cardiovascular outcome data of two meta-analyses have also suggested that rosigli- for pioglitazone are reassuring, there is a need for ran- tazone might be associated with an increased risk of domized clinical trials to be conducted in order to acute MI and death, although the majority of clin- define the appropriate role of TZDs for the treatment ical trials with TZDs were limited to subjects aged of type 2 diabetes, particularly in older people who are <65 years [173, 176–178]. A recent study examined at increased risk of CVD. the association between TZD therapy and CHF, acute To summarize: MI and mortality in older diabetics (aged ≥66 years) compared to treatment with other oral hypoglycaemic • There is no clear evidence to suggest that insulin or agents. In this population-based study of older pa- sulphonylureas are cardiotoxic tients with diabetes, TZD treatment – primarily with rosiglitazone – was associated with an increased risk • Metformin therapy seems to be cardioprotective of CHF, acute MI and mortality when compared with • The risk of metformin-induced lactic acidosis in HF other combination oral hypoglycaemic agents. During does not appear to be high a median follow up of 3.8 years, treatment with TZD monotherapy was associated with a significantly in- • Rosiglitazone is associated with an increased risk of creased risk of CHF (RR 1.60, 95% CI: 1.21–2.10, MI, HF and mortality in older diabetics 66 CH 05 CORONARY HEART DISEASE

Table 5.1 Cardiovascular event rates for combined trials. Number (%) Pioglitazone Control Hazard ratio P-value (n = 8554) (n = 7836) (95% CI) Death/myocardial infarction/stroke 375 (4.38) 450 (5.74) 0.82 (0.72–0.94) 0.005 Death 209 (2.44) 224 (2.86) 0.92 (0.76–1.11) 0.38 Myocardial infarction 131 (1.53) 159 (2.03) 0.81 (0.64–1.02) 0.08 Death/myocardial infarction 309 (3.61) 357 (4.56) 0.85 (0.73–0.99) 0.04 Stroke 104 (1.22) 131 (1.67) 0.80 (0.62–1.04) 0.09 Serious heart failure 200 (2.34) 139 (1.77) 1.41 (1.14–1.76) 0.002 Death/serious heart failure 361 (4.22) 321 (4.10) 1.11 (0.96–1.29) 0.17 Death/myocardial infarction/stroke/serious heart failure 508 (5.94) 523 (6.67) 0.96 (0.85–1.09) 0.54

Reproduced with permission from Lincoff A M, Wolski K, Nicholls S J, et al., JAMA 2007; 298: 1180–8.

• Pioglitazone reduces the risks of MI and mortality, when caring for older diabetics and, as not all will be but increases the risk of HF suited to an aggressive risk reduction, treatment must be considered on an individual basis. Any subsequent • There is a need for additional clinical trials to deﬁne individual care plan must involve not only the patients the role of thiazolidinediones in older diabetics but also their families and carers. In particular, will be necessary to recognize a patient’s diversity of beliefs 5.8 Conclusions and their attitude towards health care.

Both, ageing and diabetes have a significant impact on the cardiovascular system, increasing the risk for References developing CHD and HF in older people with dia- 1. Kannel WB, McGee DL. Diabetes and cardiovascular betes. Moreover, the size of the problem is likely to disease: the Framingham study. JAMA 1979; 241: expand as the ageing population and incidence of dia- 2035–8. betes continue to increase. The baseline risks for CVD 2. Haffner SM, Lehto S, Ronnemaa¨ T, et al. Mortality are increased in diabetics and the elderly in comparison from coronary heart disease in subjects with type 2 to non-diabetics and younger people, respectively. As diabetes and in nondiabetic subjects with and without the combination of diabetes and old age places older prior myocardial infarction. N Engl J Med. 1998; 339: diabetics at the highest baseline risk for CVD, elderly 229–34. diabetics stand to gain the most benefit from cardio- 3. Franklin SS, Gustin W, IV, Wong ND, et al.Hemo- vascular risk reduction. Medications for such reduction dynamic patterns of age related changes in blood and the treatment of CHD and HF have similar effi- pressure. The Framingham Heart Study. Circulation cacies in diabetics and non-diabetics. Older diabetics 1997; 96: 308–15. are likely to have multiple risk factors at their first 4. Blacher J, Asmar R, Djane S, et al. Aortic pulse wave presentation with diabetes, and consequently a multi- velocity as a marker of cardiovascular risk in hyper- factorial and comprehensive approach is vital to their tensive patients. Hypertension 1999; 33: 1111–17. management. Although most clinical trials have either 5. Nilsson PM, Lurbe E, Laurent S. The early life origins excluded or included few older people, there is now of vascular ageing and cardiovascular risk: the EVA syndrome J Hypertens 2008; 26: 1049–57. sufficient evidence available to suggest that the ag- 6. Samani NJ, Harst P. Biological ageing and cardiovas- gressive treatment of risk factors in this age group cular disease. Heart 2008; 94: 537–9. is both beneficial and cost effective. Although many 7. Capell BC, Collins FS, Nabel EG. Mechanisms of car- older diabetics may not achieve the recommended tar- diovascular disease in accelerated ageing syndromes. gets for risk factor reduction, for a variety of reasons Circ Res. 2007; 101: 13–26. (e.g., multiple comorbidities, polypharmacy and intol- 8. Lyons D, Roy S, Patel M, et al. Impaired nitric erance of higher doses or multiple medications), any oxide-mediated vasodilatation and total body nitric reduction in the risk factors will be beneficial. Today, oxide production in healthy old age. Clin Sci 1997; the patient’s quality of life is seen as the primary target 93: 519–22. 5.8 CONCLUSIONS 67

9. Wynne H A, Schofield S. (1996) Drug induced or- 23. Varas-Lorenzo C, Rueda de Castro A, Maguire A, et thostatic hypotension, in: R. A. Kenny (ed.), Syncope al. Prevalence of glucose metabolism abnormalities in the older patient: causes, investigations and con- and cardiovascular co-morbidity in the US elderly sequences of syncope and falls. Chapman & Hall, adult population. Pharmacoepidemiol Drug Saf 2006; London, pp. 137–54. 15: 317–26. 10. Lakatta EG. Changes in cardiovascular function with 24. Burchfiel CM, Reed DM, Marcus EB, et al. Associa- ageing. Eur Heart J 1990; 11 (Suppl. C): 22–9. tion of diabetes mellitus with coronary atherosclerosis 11. Haider AW, Larson MG, Benjamin EJ, et al. Increased and myocardial lesions. An autopsy study from the left ventricular mass and hypertrophy are associated Honolulu Heart Program. Am J Epidemiol 1993; 137: with increased risk for sudden death. J Am Coll 1328–40. Cardiol. 1998; 32: 1454–9. 25. Goraya TY, Leibson CL, Palumbo PJ, et al. 12. Lakatta EG, Levy D. Arterial and cardiac ageing: ma- Coronary atherosclerosis in diabetes mellitus. A jor shareholders in cardiovascular disease enterprises: population-based autopsy study. J Am Coll Cardiol Part I: Ageing arteries: a “set up” for vascular disease. 2002; 40: 946–53. Circulation 2003; 107: 139–46. 26. Grundy SM, Howard BV, Smith SC, Jr, et al.Pre- 13. McVeigh GE, Alen PB, Morgan DR, et al. Nitric oxide vention Conference VI: Diabetes and cardiovascular modulation of blood vessel tone identified by arterial disease. Executive summary. Conference proceeding wave form analysis. Clin Sci 2001; 100: 387–93. for healthcare professionals from a special writing 14. Cameron JD, Pinto E, Bulpitt CJ, et al. The ageing of group of the American Heart Association. Circulation elastic and muscular arteries: a comparison of diabetic 2002; 105: 2231–9. and non-diabetic subjects. Diabetes Care 2003; 26: 27. Nesto RW. Screening for asymptomatic coronary 2127–32. artery disease in diabetes. Diabetes Care 1999; 22: 1393–5. 15. Pinto E, Mensah R, Meeran K, et al. Peripheral arterial 28. Sloan FA, Bethel MA, Ruiz D, Jr, et al. The growing compliance differs between races-comparison among burden of diabetes mellitus in the US elderly popula- Asian, Afro-Caribbeans, and white Caucasians with tion. Arch Intern Med 2008; 168: 192–9. type 2 diabetes. Diabetes Care 2005; 28: 496. 29. Expert Panel on Detection, Evaluation, and Treatment 16. Ezzat VA, Duncan ER, Wheatcroft SB, et al.Therole of High Blood Cholesterol in Adults: Executive Sum- of IGF-I and its binding proteins in the development mary of The Third Report of The National Choles- of type 2 diabetes and cardiovascular disease. Dia- terol Education Program (NCEP) Expert Panel on betes, Obesity and Metabolism 2008; 10: 198–211. Detection, Evaluation, and Treatment of High Blood 17. Lee M, Gardin JM, Lynch JC, et al. Diabetes melli- Cholesterol in Adults (Adult Treatment Panel III). tus and echocardiographic left ventricular function in JAMA 2001; 285: 2486–97. free-living elderly men and women: The Cardiovas- 30. Canadian Diabetes Association. Clinical Practice cular Health Study Am Heart J 1997; 133: 36–43. Guidelines for the Prevention and Management of 18. Ren J, Sowers JR, Walsh MF and Brown RA. Reduced Diabetes in Canada. Can J Diabetes 2003; 27(Suppl. contractile response to insulin and IGF-1 in ventricu- 2): S1–S152. lar myocytes from genetically obese Zucker rats. Am 31. Colagiuri S and Best J. Lipid-lowering therapy in J Physiol. 2000; 279: H1708–H14. people with type 2 diabetes. Curr Opin Lipidol. 2002; 19. Casis O, Gallego M, Iriarte M, et al. Effects of dia- 13: 617–23. betic cardiomyopathy on regional electrophysiologic 32. Pearte CA, Furberg CD, O’Meara ES, et al. Charac- characteristics of rat ventricle. Diabetologia 2000; 43: teristics and baseline clinical predictors of future fatal 101–9. versus nonfatal coronary heart disease events in older 20. Henry, R. M. A., Paulus, W. J., Kamp, O., et al. adults. The Cardiovascular Health Study. Circulation Deteriorating glucose tolerance status is associated 2006; 113: 2177–85. with left ventricular dysfunction - The Hoorn study. 33. Gaede P, Vedel P, Larsen N, et al. Multifactorial Netherlands Journal of Medicine 2008; 66: 110–17. intervention and cardiovascular disease in patients 21. Zimmet P, Alberti KG and Shaw J. Global and so- with type 2 diabetes. N Engl J Med 2003; 348: cietal implications of the diabetes epidemic. Nature 383–93. 2001; 414: 782–7. 34. Sirois C, Moisan J, Poirier P, et al. Underuse of 22. Grundy SM, Benjamin IJ, Burke GL, et al. Diabetes cardioprotective treatment by the elderly with type 2 and cardiovascular disease: a statement for healthcare diabetes. Diabetes and Metabolism 2008; 34: 169–76. professionals from the American Heart Association. 35. Rea TD, Heckbert SR, Kaplan, RC, et al. Smoking Circulation 1999; 100: 1134–46. status and risk for recurrent coronary events after 68 CH 05 CORONARY HEART DISEASE

myocardial infarction. Ann Intern Med. 2002; 137: 48. Isommaa B, Almgren P, Tuomi T, et al. Cardio- 494–8. vascular morbidity and mortality associated with the 36. Stewart KJ. Exercise training and the cardiovascular metabolic syndrome. Diabetes Care 2001; 24: 683–9. consequences of type 2 diabetes and hypertension: 49. Lindblad U, Langer RD, Wingard DL, et al. Metabolic plausible mechanisms for improving cardiovascular syndrome and ischaemic heart disease in elderly men health. JAMA 2002; 288: 1622–31. and women. Am J Epidemiol 2001; 153: 481–9. 37. World Health Organization. (1999) Definition, Diag- 50. Scuteri A, Najjar SS, Muller DC, et al. Metabolic nosis and Classification of Diabetes Mellitus and its syndrome amplifies the age associated increases in Complications Report of a WHO Consultation. Part 1: vascular thickness and stiffness. J Am Coll Cardiol Diagnosis and Classification of Diabetes and Mellitus. 2004; 43: 1388–95. World Health Organization, Geneva. 51. Bonora E, Kiechl S, Willeit J, et al. Carotid 38. Balkau B and Charles MA. Comment on the provi- atherosclerosis and coronary heart disease in the sional report from the WHO consultation. European metabolic syndrome: prospective data from the Group for the Study of Insulin Resistance (EGIR). Bruneck Study. Diabetes Care 2003; 26: 1251–7. Diabet Med 1999; 16: 442–3. 52. Wang J, Ruotsalainen S, Moilanen L, et al.The 39. Grundy SM, Cleeman JI, Daniels SR, et al.Diag- metabolic syndrome predicts cardiovascular mortality: nosis and management of the metabolic syndrome: a 13-year follow up study in elderly non-diabetic an American Heart Association/National Heart, Lung, Finns. Euro Heart J 2007, 28: 857–64. and Blood Institute Scientific Statement. Circulation 53. Sattar N, McConnachie A, Shaper AG, et al.Can 2005; 112: 2735–52. metabolic syndrome usefully predict cardiovascu- 40. Einhorn D, Reaven GM, Cobin RH, et al. American lar disease and diabetes? Outcome data from two prospective studies. Lancet 2008; 371: 1927–35. College of Endocrinology position statement on the 54. Casiglia E, Mazza A, Tikhonoff V, et al. Total choles- insulin resistance syndrome. Endocr Pract 2003; 9: terol and mortality in the elderly. J Intern Med 2003; 237–52. 254: 353–62. 41. Alberti KG, Zimmet P and Shaw J. The metabolic 55. Volpato S, Zuliani G, Guralnik JM, et al. The inverse syndrome – a new worldwide definition. Lancet 2005; association between age and cholesterol levels among 366: 1059–62. older patients: the role of poor health status. Geron- 42. Ravera M, Ratto E, Vettoretti S, et al. Microalbu- tology 2001; 47: 36–45. minuria and subclinical cerebrovascular damage in 56. Psaty BM, Anderson M, Kronmal RA, et al.Theas- essential hypertension. J Nephrol 2002; 15: 519–24. sociation between lipid levels and the risks of incident 43. Scuteri A, Najjar SS, Morrell CH, et al. The metabolic myocardial infarction, stroke, and total mortality: the syndrome in older individuals: prevalence and predic- Cardiovascular Health Study. J Am Geriatr Soc. 2004; tion of cardiovascular events. Diabetes Care 2005; 28: 52: 1639–47. 882–7. 57. Aronow WS and Ahn C. Incidence of new coro- 44. Empana JP, Zureik M, Gariepy J, et al. The metabolic nary events in older persons with prior myocardial syndrome and the carotid artery structure in noninsti- infarction and serum low-density lipoprotein choles- tutionalized elderly subjects. The Three-City Study. terol ≥125 mg/dl treated with statins versus no Stroke 2007; 38: 893–9. lipid-lowering drug. Am J Cardiol. 2002; 89: 67–9. 45. Hildrum B, Mykletun A, Hole T, et al. Age-specific 58. Aronow WS, Ahn C and Gutstein H. Incidence prevalence of the metabolic syndrome defined by of new atherothrombotic brain infarction in older the International Diabetes Federation and the Na- persons with prior myocardial infarction and serum tional Cholesterol Education Program: the Norwegian low-density lipoprotein cholesterol ≥125 mg/dl HUNT 2 study. BMC Public Health 2007, 7: 220. treated with statins versus no lipid-lowering drug. J 46. Maggi S, Noale M, Gallina P, et al. Metabolic syn- Gerontol A Biol Sci Med Sci. 2002; 57: M333–5. drome, diabetes, and cardiovascular disease in an el- 59. Foody JM, Rathore SS, Galusha D, et al. derly Caucasian cohort: the Italian longitudinal study Hydroxymethylglutaryl-CoA reductase inhibitors on ageing. J Gerontol 2006; 61: 505–10. in older persons with acute myocardial infarction: 47. Alexander CM, Landsman PB, Teutsch SM, et evidence for an age statin interaction. J Am Geriatr al. NCEP-defined metabolic syndrome, diabetes, Soc. 2006; 54: 421–30. and prevalence of coronary heart disease among 60. Anum EA and Adera T. Hypercholesterolemia and NHANES III participants age 50 years and older. coronary heart disease in the elderly: a meta-analysis. Diabetes 2003; 52: 1210–14. Ann Epidemiol. 2004; 14: 705–21. 5.8 CONCLUSIONS 69

61. Baigent C, Keech A, Kearney PM, et al. The Choles- Treatment (HOT) randomised trial. Lancet 1998; 351: terol Treatment Trialists’ (CTT) Collaborators: Ef- 1755–62. ficacy and safety of cholesterol-lowering treatment: 73. Tuomilehto J, Rastenyte D, Birkenhager WH, et al. prospective meta-analysis of data from 90,056 partic- Systolic Hypertension in Europe Trial Investigators. ipants in 14 randomised trials of statins. Lancet 2005; Effects of calcium-channel blockade in older patients 366: 1267–78. with diabetes and systolic hypertension. N Engl J 62. Heart Protection Study Collaborative Group (2002) Med. 1999; 340: 677–84. Heart Protection Study of cholesterol lowering with 74. Curb JD, Pressel SL, Cutler JA, et al. Systolic Hyper- simvastatin in 20 536 high-risk individuals: a ran- tension in the Elderly Program Cooperative Research domised placebo-controlled trial. Lancet, 360, 7–22. Group. Effect of diuretic based antihypertensive treat- 63. Heart Protection Study Collaborative Group (2003) ment on cardiovascular disease risk in older diabetic Heart Protection Study of cholesterol lowering patients with isolated systolic hypertension. JAMA. with simvastatin in 5963 people with diabetes: a 1996; 276: 1886–92. randomised placebo-controlled trial. Lancet, 361, 75. McFarlane SI, Jacober SJ, Winer N, et al. Control of 2005–16. cardiovascular risk factors in patients with diabetes 64. Costa J, Borges M, David C, et al. Efficacy of lipid and hypertension at urban academic medical centers. lowering drug treatment for diabetic and non-diabetic Diabetes Care 2002; 25: 718–23. patients: meta-analysis of randomised controlled tri- 76. Gianni M, Bosch J, Pogue J, et al. Effect of long-term als. Br Med J 2006; 332; 1115–24. ACE-inhibitor therapy in elderly vascular disease pa- 65. Neil HAW, DeMicco DA, Luo D, et al.Analysisof tients Eur Heart J 2007; 28: 1382–8. efficacy and safety in patients aged 65–75 years at 77. Ravid M, Brosh D, Levi Z, et al. Use of enalapril to attenuate decline in renal function in normotensive, randomisation. Diabetes Care 2006; 29: 2378–84. normoalbuminuric patients with type 2 diabetes mel- 66. Mihaylova B, Briggs A, Armitage J, et al.The litus. A randomized, controlled trial. Ann Intern Med. Heart Protection Study Collaborative Group: 1998; 128: 982–8. Cost-effectiveness of simvastatin in people at 78. Ahmad J, Siddiqui MA and Ahmad H. Effective post- different levels of vascular risk: economic analysis ponement of diabetic nephropathy with enalapril in of a randomised trial in 20,536 individuals. Lancet normotensive type 2 diabetic patients with microal- 2005; 365: 1779–85. buminuria. Diabetes Care 1997; 20: 1576–81. 67. Sowers JR, Epstein M, Frohlich ED. Diabetes, hyper- 79. Bangalore S, Parkar S, Grossman E, et al.A tension, and cardiovascular disease: an update. Hy- meta-analysis of 94,492 patients with hypertension pertension 2001; 37: 1053–9. treated with beta blockers to determine the risk of 68. Cowie, C. C. and Harris, M. I. (1995) Physical and new-onset diabetes mellitus. Am J Cardiol 2007; Metabolic Characteristics of Persons with Diabetes. 100: 1254–62. 2nd edition. National Institutes of Health, National 80. Elliott WJ and Meyer PM. Incident diabetes in Institute of Diabetes and Digestive and Kidney Dis- clinical trials of antihypertensive drugs: a network eases, Bethesda, MD, pp. 117–64. meta-analysis. Lancet 2007; 369: 201–7. 69. Gress TW, Nieto FJ, Shahar E, et al. Hypertension and 81. Lithell H, Pollare T and Vessby B. Metabolic ef- antihypertensive therapy as risk factors for type 2 di- fects of pindolol and propranolol in a double-blind abetes mellitus: Atherosclerosis Risk in Communities cross-over study in hypertensive patients. Blood Press Study. N Engl J Med. 2000; 342: 905–12. 1992; 1: 92–101. 70. Chobanian AV, Bakris GL, Black HR, et al.The 82. Bakris GL, Fonseca V, Katholi RE, et al. Metabolic Seventh Report of the Joint National Committee on effects of carvedilol vs metoprolol in patients with Prevention, Detection, Evaluation, and Treatment of type 2 diabetes mellitus and hypertension: a random- High Blood Pressure: the JNC 7 Report. JAMA. 2003; ized controlled trial. JAMA 2004; 292: 2227–36. 289: 2560–71. 83. ALLHAT Officers and Coordinators for the ALLHAT 71. UK Prospective Diabetes Study Group. Tight blood Collaborative Research Group. (2002) Major out- pressure control and risk of macrovascular and mi- comes in high-risk hypertensive patients randomized crovascular complications in type 2 diabetes: UKPDS to angiotensin converting enzyme inhibitor or calcium 38. Br Med J. 1998; 317: 703–13. channel blocker vs diuretic: The Antihypertensive and 72. Hansson L, Zanchetti A, Carruthers SG, et al.andthe Lipid-Lowering Treatment to Prevent Heart Attack HOT Study Group. Effects of intensive blood pressure Trial (ALLHAT). JAMA, 288, 2981–97. lowering and low dose aspirin in patients with hyper- 84. Brenner BM, Cooper ME, de Zeeuw D, et al. Effects tension: principal results of the Hypertension Optimal of losartan on renal and cardiovascular outcomes in 70 CH 05 CORONARY HEART DISEASE

patients with type 2 diabetes and nephropathy. N Engl 96. DECODE Study Group. (2001) European Diabetes J Med. 2001; 345: 861–9. Epidemiology Group: Glucose tolerance and cardio- 85. Parving HH, Lehnert H, Brochner-Mortensen J, et vascular mortality: comparison of fasting and 2-hour al. The effect of irbesartan on the development of diagnostic criteria. Arch Intern Med., 161: 397–405. diabetic nephropathy in patients with type 2 diabetes. 97. Selvin E, Marinopoulos S, Berkenblit G, et al. N Engl J Med. 2001; 345: 870–8. Meta-analysis: glycosylated hemoglobin and 86. Lewis EJ, Hunsicker LG, Clarke WR, et al. Renopro- cardiovascular disease in diabetes mellitus. Ann tective effect of the angiotensin receptor antagonist Intern Med 2004; 141: 421–31. irbesartan in patients with nephropathy due to type 2 98. Milicevic Z, Raz I, Beattie SD, et al. Natural history diabetes. N Engl J Med. 2001; 345: 851–60. of cardiovascular disease in patients with diabetes. 87. Lindholm LH, Ibsen H, Dahlof¨ B, et al. Cardiovascu- Diabetes Care 2008; 31 (Suppl. 2): S155–60. lar morbidity and mortality in patients with diabetes 99. Hanefeld M, Cagatay M, Petrowitsch T, et al. Acar- in the Losartan Intervention For Endpoint reduction in bose reduces the risk for myocardial infarction in type hypertension study (LIFE): a randomised trial against 2 diabetic patients: meta-analysis of seven long-term atenolol. Lancet 2002; 359: 1004–10. studies. Eur Heart J 2004; 25: 10–16. 88. Matchar DB, McCrory DC, Orlando LA, et al. 100. Carmassi F, Morale M and Puccetti R. Coagula- Systematic Review: Comparative effectiveness tion and ﬁbrinolytic system impairment in insulin of angiotensin converting enzyme inhibitors and dependent diabetes mellitus. Thromb Res 1992; 67: angiotensin II receptor blockers for treating essential 643–54. hypertension. Ann Intern Med. 2008; 148: 16–29. 101. Walsh MF, Dominguez LJ and Sowers JR. Metabolic 89. Beckett NS, Peters R, Fletcher AE, et al. Treatment abnormalities in cardiac ischaemia. Cardiol Clin 1995; of hypertension in patients 80 years of age or older. 13: 529–38. N Engl J Med 2008; 358: 1887–98. 102. Beckman JA, Creager MA and Libby P. Diabetes and 90. Langford HG, Cutter G, Oberman A, et al.The atherosclerosis. JAMA 2002; 287: 2570–81. effect of thiazide therapy on glucose, insulin and 103. Gupta OP and Phatak S. Pandemic trends in preva- cholesterol metabolism and of glucose on potassium: lence of diabetes mellitus and associated coronary results of a cross-sectional study in patients from the heart disease in India - their causes and prevention. Hypertension Detection and Follow-up Program. J Int J Diab 2003; 23: 37–50. Hum Hypertens. 1990; 4: 491–500. 104. Kodali VRR. (2002) Epidemiology of diabetes melli- 91. Morigi M, Angioletti S, Imberti B, et al. tus in migrant populations, in: M. M. S. Ahuja, B. B. Leukocyte-endothelial interaction is augmented by Tripathi,S.G.P.Moses,H.B.Chandalia,A.K.Das, high glucose concentrations and hyperglycemia in a P. V. Rao and S. V. Madhu (eds), RSSDI Textbook NF-kB dependent fashion. J Clin Invest 1998; 101: of Diabetes Mellitus, National Book Depot, India, pp. 1905–15. 113–19. 92. Turner RC, Millns H, Neil HAW, et al. Risk factors 105. Mohan V, Deepa R, Shanthirani S, et al. Prevalence for coronary artery disease in non-insulin dependent of coronary artery disease and its relationship to lipids diabetes mellitus: United Kingdom prospective dia- in a selected population in south India. J. Am. Coll betes study (UKPDS:23). Br Med J 1998; 316: 823–8. Cardiol 2001; 38: 687–97. 93. Letho S, Ronnemaa T, Pyorala K, et al. Poor gly- 106. Mohan V, Deepa R, Harath SP, et al. Lipoprotein caemic control predicts coronary heart disease events (a) is an independent risk factor for coronary artery in patients with type 1 diabetes without nephropathy. disease in NIDDIM patients in south India. Diabetes Arterioscler Thromb Vasc Biol 1999; 13: 1014–19. Care, 1998; 21: 1819–23. 94. Turner RC, Cull CA, Frighi V, et al. Glycemic con- 107. Saaddine JB, Engelgau MM, Beckles GL, et al.A trol with diet, sulfonylurea, metformin, or insulin in diabetes report card for the United States: quality patients with type 2 diabetes mellitus: progressive re- of care in the 1990s. Ann Intern Med 2002; 136: quirement for multiple therapies (UKPDS 49): UK 565–74. Prospective Diabetes Study (UKPDS) Group. JAMA 108. Kirk JK, Bell RA, Bertoni AG, et al. Qualitative 1999; 281: 2005–12. review of studies of diabetes preventive care among 95. DECODE Study Group. (2003) European Diabetes minority patients in the United States, 1993–2003 Epidemiology Group: Is the current deﬁnition for (Review). Am J Manag Care 2005; 11: 349–60. diabetes relevant to mortality risk from all causes 109. Ness J, Aronow WS. Prevalence of coronary and cardiovascular and noncardiovascular diseases? artery disease, ischemic stroke, peripheral arterial Diabetes Care, 26: 688–96. disease, and coronary revascularization in older 5.8 CONCLUSIONS 71

African-Americans, Asians, Hispanics, whites, men, cardiovascular disease. 10 year follow-up of the and women. Am J Cardiol 1999; 84: 932–3. Hoorn Study. Eur Heart J. 2003; 24: 1406–13. 110. Gillum RF, Mussolino ME and Madans JH. Dia- 123. Kronmal RA, Barzilay JI, Smith NL, et al. Mortality betes mellitus, coronary heart disease incidence, and in pharmacologically treated older adults with dia- death from all causes in African American and Euro- betes: the Cardiovascular Health Study 1989–2001. pean American women: the NHANES I epidemiologic PLoS Med. 2006; 3: e400. follow-up study. J Clin Epidemiol 2000; 53: 511–18. 124. Fox CS, Sullivan L, D’Agostino RB, Sr, et al.The 111. Karter AJ, Ferrara A, Liu JY, et al. Ethnic disparities significant effect of diabetes duration on coronary in diabetic complications in an insured population. heart disease mortality: the Framingham Heart Study. JAMA 2002; 287: 2519–27. Diabetes Care 2004; 27: 704–8. 112. Young BA, Maynard C and Boyko EJ. Racial differ- 125. Hu FB, Stampfer MJ, Haffner SM, et al. Elevated risk ences in diabetic nephropathy, cardiovascular disease, of cardiovascular disease prior to clinical diagnosis of and mortality in a national population of veterans. Di- type 2 diabetes. Diabetes Care 2002; 25: 1129–34. abetes Care 2003; 26: 2392–9. 126. Chen J, Marciniak TA, Radford MJ, et al. 113. Chaturvedi N, Jarrett J, Morrish N, et al. Differences Beta-blocker therapy for secondary prevention of in mortality and morbidity in African Caribbean and myocardial infarction in elderly diabetic patients. European people with non-insulin dependent diabetes Results from the national cooperative cardiovascular mellitus. Br Med J 1996; 313: 848–52. project. J Am Coll Cardiol 1999; 34: 1388–94. 114. Bertoni AG, Kirk JK, Case LD, et al. The effects of 127. Chyun D, Vaccarino V, Murillo J, et al. Cardiac out- race and region on cardiovascular morbidity among comes after myocardial infarction in elderly patients elderly Americans with diabetes. Diabetes Care 2005; with diabetes mellitus. Am J Critical Care 2002; 11: 28: 2620–5. 504–19. 115. Fein F and Scheuer J. (1990) Heart disease in dia- 128. Kosiborod M, Rathore SS, Inzucchi SE, et al.Admis- betes mellitus: theory and practice. In: H. Rifkin and sion glucose and mortality in elderly patients hospital- D. Porte, Jr (eds), Ellenberg and Rifkin’s Diabetes ized with acute myocardial infarction implications for Mellitus: Theory and Practice. 4th edition. Elsevier patients with and without recognized diabetes. Circu- Science, Inc., New York, pp. 812–23. lation 2005; 111: 3078–86. 116. Miettinen H, Lehto S, Salomaa V, et al. Impact 129. Mehta RH, Ruane TJ, McCargar PA, et al. The treat- of diabetes on mortality after the first myocardial ment of elderly diabetic patients with acute myocar- infarction. Diabetes Care 1998; 21: 69–75. dial infarction insight from Michigan’s Cooperative 117. Booth GL, Kapral MK, Fung K, et al. Relation be- Cardiovascular Project. Arch Intern Med. 2000; 160: tween age and cardiovascular disease in men and 1301–6. women with diabetes compared with non-diabetic 130. Donahue RP, Goldberg RJ, Chen Z, et al.Thein- people: a population-based retrospective cohort study. fluence of sex and diabetes mellitus on survival fol- Lancet 2006; 368: 29–36. lowing acute myocardial infarction: a communitywide 118. Chiariello M, Indolfi C, Cotecchia MR, et al.Asymp- perspective. J Clin Epidemiol. 1993; 46: 245–52. tomatic transient changes during ambulatory ECG 131. Woodfield S, Ludergan C, Reiner J, et al. Angio- monitoring in diabetic patients. Am Heart J. 1985; graphic findings and outcome in diabetic patients 110: 529–34. treated with thrombolytic therapy for acute myocar- 119. Marshall SM and Flyvbjerg A. Prevention and early dial infarction: the GUSTO-I experience. J Am Coll detection of vascular complications of diabetes. Br Cardiol. 1996; 28: 1661–9. Med J 2006; 333: 475–80. 132. Malmberg K, Ryden L, Efendic S, et al. Randomized 120. Lee WL, Cheung AM, Cape D, et al. Impact of trial of insulin-glucose infusion followed by subcu- diabetes on coronary artery disease in women and taneous insulin treatment in diabetic patients with men: a meta-analysis of prospective studies. Diabetes acute myocardial infarction (DIGAMI study): effects Care 2000; 23: 962–8. on mortality at one year. J Am Coll Cardiol. 1995; 121. Huxley R, Barzi F and Woodward M. Excess risk of 26: 57–65. fatal coronary heart disease associated with diabetes 133. Malmberg K. DIGAMI (Diabetes Mellitus, Insulin in men and women: meta-analysis of 37 prospective Glucose Infusion in Acute Myocardial Infarction) cohort studies. Br Med J 2006; 332: 73–8. Study Group. Prospective randomised study of inten- 122. Becker A, Bos G, de Vegt F, et al. Cardiovas- sive insulin treatment on long term survival after acute cular events in type 2 diabetes: comparison with myocardial infarction in patients with diabetes melli- nondiabetic individuals without and with prior tus. Br Med J. 1997; 314: 1512–15. 72 CH 05 CORONARY HEART DISEASE

134. Malmberg K, Ryden L, Wedel H, et al. DIGAMI 2 of complications in patients with type 2 diabetes Investigators. Intense metabolic control by means of (UKPDS 33). Lancet 1998; 352: 837–53. insulin in patients with diabetes mellitus and acute 148. Heart Outcomes Prevention Evaluation (HOPE) myocardial infarction (DIGAMI 2): effects on mor- Study Investigators. Effects of ramipril on cardio- tality and morbidity. Eur Heart J. 2005; 26: 650–61. vascular and microvascular outcomes in people with 135. Fath-Ordoubadi F and Beatt KJ. Glucose-insulin- diabetes mellitus: results of the HOPE study and potassium therapy for treatment of acute myocar- MICRO-HOPE substudy. Lancet 2000; 355: 253–9. dial infarction: an overview of randomized placebo- 149. Shepherd J, Barter P, Carmena R, et al. for the Treat- controlled trials. Circulation 1997; 96: 1152–6. ing to New Targets Investigators. (2006) Effect of 136. Kloner RA and Nesto RW. Glucose-insulin-potassium lowering LDL cholesterol substantially below cur- for acute myocardial infarction continuing contro- rently recommended levels in patients with coronary versy over cardioprotection. Circulation 2008; 117: heart disease and diabetes: the Treating to New Tar- 2523–33. gets (TNT) study. Diabetes Care, 29, 1220–6. 137. Nichols GA, Gullion CM, Koro CE, et al.Theinci- 150. Bertoni AG, Hundley WG, Massing MW, et al. dence of congestive heart failure in type 2 diabetes: Heart failure prevalence, incidence, and mortality in an update. Diabetes Care 2004; 27: 1879–84. the elderly with diabetes. Diabetes Care 2004; 27: 138. Carr AA, Kowey PR, Devereux RB, et al. Hospi- 699–703. talisations for new heart failure among subjects with 151. Vaur L, Gueret P, Lievre M, et al. Development of diabetes mellitus in the RENAAL and LIFE studies. congestive heart failure in type 2 diabetic patients with Am J Cardiol 2005; 96: 1530–6. microalbuminuria or proteinuria: observations from 139. Stratton IM, Adler AI, Neil HA, et al. Association the DIABHYCAR (type 2 DIABetes Hypertension CArdiovascular Events and Ramipril) study. Diabetes of glycaemia with macrovascular and microvascu- Care 2003; 26: 855–60. lar complications of type-2 diabetes (UKPDS 35): 152. Epshteyn V, Morrison K, Krishnaswamy P, et al. prospective observational study. Br Med J 2000; 321: Utility of B-Type Natriuretic Peptide (BNP) as a 405–12. screen for left ventricular dysfunction in patients with 140. Wong TY, Rosamond W, Chang PP, et al. Retinopa- diabetes. Diabetes Care 2003; 26: 2081–7. thy and risk of congestive heart failure. JAMA 2005; 153. Banerjee P, Clark AL, Nikitin N, et al. Diastolic heart 293: 63–9. failure: paroxysmal or chronic? Eur J Heart Fail 2004; 141. Yip GWK, Ho PPY, Woo KS, et al. Comparison of 6: 427–31. frequencies of left ventricular systolic and diastolic 154. Haas SJ, Vos T, Gilbert RE, et al. Are beta-blockers heart failure in Chinese living in Hong Kong. Am J as efficacious in patients with diabetes mellitus as in Cardiol 1999; 84: 563–7. patients without diabetes mellitus who have chronic 142. Bell DS. Heart failure: the frequent, forgotten, and heart failure? A meta-analysis of large-scale clinical often fatal complication of diabetes. Diabetes Care trials. Am Heart J 2003; 146: 848–53. 2003; 26: 2433–41. 155. Deedwania PC, Giles TD, Klibaner M, et al.Effi- 143. Heart Outcomes Prevention Evaluation Study Inves- cacy, safety and tolerability of metoprolol CR/XL in tigators: Effects of ramipril on cardiovascular and patients with diabetes and chronic heart failure: ex- microvascular outcomes in people with diabetes mel- periences from MERIT-HF. Am Heart J 2005; 149: litus: results of the HOPE study and MICROHOPE 159–67. substudy. Lancet 2000; 355: 253–9. 156. Wlodarczyk JH, Keogh A, Smith K, et al. CHART: 144. Nichols GA, Hillier TA, Erbey JR, et al. Congestive congestive cardiac failure in hospitals, an Australian heart failure in type 2 diabetes: prevalence, incidence, review of treatment. Heart Lung Circ 2003; 12: and risk factors. Diabetes Care 2001; 24: 1614–19. 94–102. 145. Davis RC, Hobbs FD, Kenkre JE, et al. Prevalence of 157. Del Sindaco DA, Pulignano GB, Cioffi GC, et al. left ventricular systolic dysfunction and heart failure Safety and efficacy of carvedilol in very elderly di- in high risk patients: community based epidemiolog- abetic patients with heart failure. J Cardiovasc Med ical study. Br Med J 2002; 325: 1156. 2007; 8: 675–82. 146. Aronow WS and Ahn C. Incidence of heart failure 158. Shekelle PG, Rich MW, Morton SC, et al.Effi- in 2,737 older persons with and without diabetes cacy of angiotensin converting enzyme inhibitors and mellitus. Chest 1999; 115: 867–8. beta-blockers in the management of left ventricular 147. UK Prospective Diabetes Study (UKPDS). Intensive systolic dysfunction according to race, gender, and di- blood glucose control with sulphonylureas or in- abetic status: a meta-analysis of major clinical trials. sulin compared with conventional treatment and risk J Am Coll Cardiol 2003; 41: 1529–38. 5.8 CONCLUSIONS 73

159. Young JB, Dunlap ME, Pfeffer MA, et al., for the formin on complications in overweight patients with Candesartan in Heart Failure Assessment of Reduc- type 2 diabetes (UKPDS 34). Lancet 1998; 352: tion in Mortality morbidity (CHARM) Investigators. 854–65. (2004) Mortality and morbidity reduction with can- 170. Eurich DT, Majumdar SR, McAlister FA, et al.Im- desartan in patients with chronic heart failure and proved clinical outcomes associated with metformin left ventricular systolic dysfunction: results of the in patients with diabetes and heart failure. Diabetes CHARM low-left ventricular ejection fraction trials. Care 2005; 28: 2345–51. Circulation, 110: 2618–26. 171. Kahn SE, Haffner SM, Heise MA, et al. Glycaemic 160. Cohn JN and Tognoni G. The Valsartan Heart Fail- durability of rosiglitazone, metformin, or glyburide ure Trial Investigators. A randomized trial of the monotherapy. N Engl J Med. 2006; 355: 2427–43. angiotensin-receptor blocker valsartan in chronic heart 172. Stettler C, Allemann S, Juni P, et al. Glycemic con- failure. N Engl J Med 2001; 345: 1667–75. trol and macrovascular disease in types 1 and 2 dia- 161. Fernandez HM, Leipzig RM, Larkin RJ, et al.The betes mellitus: meta-analysis of randomized trials. Am Randomized Aldactone Evaluation Study Investiga- Heart J. 2006; 152: 27–38. tors. Spironolactone in patients with heart failure. N 173. Lincoff AM, Wolski K, Nicholls SJ, et al. Pioglitazone Engl J Med 2000; 342: 132–4. and risk of cardiovascular events in patients with type 162. Huxley R, Barzi F and Woodward M. Excess risk of 2 diabetes mellitus: a metaanalysis of randomized fatal coronary heart disease associated with diabetes trials. JAMA. 2007; 298: 1180–8. in men and women: meta-analysis of 37 prospective 174. Berlie HD, Kalus JS and Jaber LA. Thiazolidine- cohort studies. Br Med J 2006; 332: 73–8. diones and the risk of edema: a meta-analysis. Di- 163. Ahmed A, Aban IB, Vaccarino V, et al.A abetes Res Clin Pract. 2007; 76: 279–89. propensity-matched study of the effect of diabetes on 175. Karalliedde J and Buckingham RE. Thiazolidine- the natural history of heart failure: variations by sex diones and their ﬂuid-related adverse effects: facts, and age Heart 2007; 93: 1584–90. ﬁction and putative management strategies. Drug 164. Klimt CR, Knatterud GI, Meinert CL, et al. University Safety 2007; 30: 741–53. Group Diabetes Program. A study of the effects of 176. Nissen SE and Wolski K. Effect of rosiglitazone hypoglycaemic agents on vascular complications in on the risk of myocardial infarction and death from patients with adult-onset diabetes, II: mortality results. cardiovascular causes. N Engl J Med. 2007; 356: Diabetes 1970; 19 (Suppl.): S789–830. 2457–71. 165. Karter AJ, Ahmed AT, Liu J, et al. Pioglitazone ini- 177. Singh S, Loke YK and Furberg CD. Long-term tiation and subsequent hospitalisation for congestive risk of cardiovascular events with rosiglitazone: a heart failure. Diabet Med 2005; 22: 986–93. meta-analysis. JAMA. 2007; 298: 1189–95. 166. Masoudi FA, Inzucchi SE, Wang Y, et al. Thiazo- 178. Singh S, Loke YK and Furberg CD. Thiazolidine- lidinediones, metformin, and outcomes in Relation diones and heart failure: a teleo-analysis. Diabetes between chronic heart failure and diabetes mellitus Care 2007; 30: 2148–53. 1237 older patients with diabetes and heart failure: an 179. Lipscombe LL, Gomes T, Le’vesque LE, et al.Thi- observational study. Circulation 2005; 111: 583–90. azolidinediones and cardiovascular outcomes in older 167. Domanski M, Krause-Steinrauf H, Deedwania, P. et patients with diabetes. JAMA 2007; 298: 2634–43. al. The effect of diabetes on outcomes of patients with 180. Derosa G, Cicero AF, D’Angelo A, et al. Effects of advanced heart failure in the BEST trial. J Am Coll 1 year of treatment with pioglitazone or rosiglitazone Cardiol 2003; 42: 914–22. added to glimepiride on lipoprotein(a) and homocys- 168. Smooke S, Horwich TB and Fonarow GC. teine concentrations in patients with type 2 diabetes Insulin-treated diabetes is associated with a marked mellitus and metabolic syndrome: a multicenter, ran- increase in mortality in patients with advanced heart domized, double-blind, controlled clinical trial. Clin failure. Am Heart J 2005; 149: 168–74. Ther. 2006; 28: 679–88. 169. UK Prospective Diabetes Study (UKPDS) Group. Effect of intensive blood-glucose control with met- 6 Stroke and Diabetes

Janice E. O’Connell and Christopher S. Gray Department of Geriatric Medicine, University of Newcastle, Sunderland Royal Hospital, Sunderland, UK

tus (T2DM) are known to be important risk factors for Key messages stroke [1, 2]. However, this association is primarily for ischaemic stroke, and there is no proven link between • Type 2 diabetes mellitus is the predominant form diabetes and primary intracerebral haemorrhage. The in stroke patients. It is not only a risk factor for predominant form of diabetes in older people is T2DM, cerebrovascular disease but also increases the and in this chapter we will focus on the relationship be- likelihood of a poor clinical outcome following tween T2DM, ischaemic stroke and transient ischaemic acute stroke. attack (TIA). • Hyperglycaemia after acute stroke is associated with a poor prognosis, both in diabetic and non-diabetic individuals. 6.1.1 Definitions • Acute treatment of post-stroke hyperglycaemia The World Health Organization (WHO) diagnostic cri- with insulin has not been shown to be beneficial teria make an arbitrary time distinction at 24 h from in recent clinical trials. symptom onset between TIAs, where transient neuro- • Primary and secondary stroke prevention in patients with type 2 diabetes requires good logical dysfunction lasts less than 24 h, and the com- diabetic control, plus management of all other pleted stroke, with symptoms and signs persisting be- vascular risk factors. yond this time [3]. However, more recent evidence shows that most TIAs resolve within 1 h and, moreover, TIA patients with symptoms lasting more than 6 h 6.1 Introduction have neuroradiological evidence of ischaemic change on magnetic resonance imaging (MRI) brain scan. A The spectrum of cerebrovascular disease ranges from new definition for TIA has therefore been proposed: reversible neurological symptoms and signs due to ‘...a brief episode of neurological dysfunction with transient cerebral ischaemia to permanent neurological symptoms lasting less than one hour and without evi- deficit consequent on brain infarction or primary in- dence of acute infarction’ [4]. tracerebral haemorrhage. Furthermore, cerebrovascular Within the group of patients with ischaemic stroke, disease is now recognized as a major cause of cognitive there are several potential pathophysiological mecha- decline, not only due to vascular dementia but also as nisms. Approximately 29–44% will be due to throm- a contributor to Alzheimer’s disease through a mixed bosis in situ, 20–25% to cardioembolism, 13–21% to vascular and Alzheimer’s-type pathology. Both type 1 small-artery disease (lacunar stroke) and 15–17% to diabetes mellitus (T1DM) and type 2 diabetes melli- mixed or undetermined aetiologies [5]. In diabetes and

Diabetes in Old Age Third Edition Edited by Alan J. Sinclair © 2009 John Wiley & Sons Ltd. ISBN: 978-0-470-06562-4 76 CH 06 STROKE AND DIABETES impaired glucose tolerance, atherosclerotic changes in 6.2 Diabetes, hyperglycaemia and the vasculature are accelerated, with a threefold increased risk for asymptomatic carotid artery stenosis stroke risk in older people seen in T2DM [6]. There remains conflicting evi- Both T1DM and T2DM are important risk factors for dence as to the clinical and pathological type of is- cerebrovascular disease, conferring a two- to three- chaemic stroke that predominates in diabetic individu- fold increased risk of first and recurrent stroke [1, als. Large-artery atherosclerotic disease was previously 2]. In older people, the prevalent form is T2DM, felt to be more severe in patients with diabetes mellitus seen in up to 7% of individuals. Glucose tolerance (DM), but more recent evidence suggested an associ- is known to decline with ageing, with estimates that ation between DM and small-vessel lacunar events [7, a further 7.7–14.8% of older adults will have unrec- 8]. A meta-analysis of population-based studies how- ognized T2DM according to 1997 American Diabetes ever, failed to demonstrate any association between Association and 1985 World Health Organisation cri- DM and small-vessel disease (OR 1.1 95% CI: 0.8–1.4, teria, respectively [15]. The prevalence of T2DM in an p = 0.82) [9]. elderly Caucasian population may be as high as 20%, with even higher rates in other ethnic groups such as 6.1.2 Epidemiology of stroke south Asians living in Britain [16]. Although most diabetic patients have multiple risk factors for vascular In both the UK and USA, cerebrovascular disease is disease, diabetes remains an independent risk factor for the third major cause of death, and the most important stroke across all age groups. Furthermore, as a marker cause of severe adult disability in the community of arteriosclerotic vascular disease, it is estimated that [10]. Stroke is also the single most expensive medical the prevalence of carotid artery disease in elderly di- disorder, consuming up to 6% of the total clinical abetics is 20% [17]. The increased likelihood of suf- budget [11]. Against this background of the high costs fering cardiovascular disease (CVD) is not confined of medical and social care, the major risk factors for to people with diabetes, but also includes those with cerebrovascular disease have been well defined and impaired glucose tolerance, asymptomatic non-fasting the management of modifiable risk factors outlined in hyperglycaemia and hyperinsulinaemia [18]. national guidelines and service frameworks [12, 13]. Since most strokes occur in older people, it follows Until recently, the burden of stroke on health and that the majority of diabetic stroke patients will have social services has been relatively under-recognized T2DM. Published estimates of the actual stroke risk in comparison with other diseases such as cancer in diabetics vary, due to differences in the popula- and coronary heart disease (CHD). However, with tions studied, the diagnostic criteria employed or the the recent publication of the Department of Health use of indirect measures such as glycated haemoglobin National Stroke Strategy comes the recognition of the (HbA1C) concentration. A fairly reliable risk estimate importance of stroke as a major cause of morbidity and for stroke risk in T2DM may be obtained from a re- mortality in the UK [14]. There are likely to be major cent cohort study of all known T2DM cases on the changes in the organization and delivery of stroke UK General Practice Research Database [19]. Here, services in England and Wales in the next decade. 41 799 patients with T2DM aged 35–89 years at outset Stroke is primarily a disorder of old age, with its were followed for 7 years, during which the absolute peak incidence seen in those over 75 years of age. Nev- rate of stroke was 11.91 per 1000 person-years, com- ertheless, a significant proportion of patients is younger pared to 5.55 per 1000 person-years for non-diabetic and has potentially modifiable risk factors, including matched controls. The age-adjusted hazard ratio for T2DM. Over the past few decades, we have witnessed stroke in T2DM compared with controls was 2.19 demographic shifts in our population towards the ex- (95% CI: 2.09–2.32). Additional stroke risk factors tremes of old age, with maximal population expansion in T2DM included the duration of diabetes, smok- in those aged over 75 years. Despite guidance-based ining, obesity, atrial fibrillation and hypertension. The terventions for the detection and early management of Nurses’ Health Study, a female-only cohort, also con- vascular risk factors, some (including T2DM) remain firmed the relationship between T2DM and ischaemic under-recognized, especially in the elderly. It seems stroke, with a relative risk of 2.3 (95% CI: 2.0–2.6) likely therefore that the incidence of stroke in our age- [20]. In this study, the risk of stroke was associ- ing population will continue to rise. ated with the duration of T2DM, and similar risks 6.2 DIABETES, HYPERGLYCAEMIA AND STROKE RISK IN OLDER PEOPLE 77 for large-artery and lacunar infarction were observed. their disease. Data from the United Kingdom Prospec- The incidence and risk factors for stroke in T2DM tive Diabetes Study (UKPDS) confirmed that over a were also examined in another recent cohort study 9-year period, 20% of individuals with T2DM would of 14 432 patients attending 201 Italian diabetes cen- experience macrovascular complications, including is- tres, which found 296 incident strokes over a 4-year chaemic stroke, whilst only 9% would have microvas- follow-up period [21]. Pre-existing cardiovascular dis- cular problems [24]. A recent inception cohort study ease was identified as an important risk factor, with an in Saskatchewan, Canada confirmed that macrovascu- age-standardized incidence of stroke in men of 13.7 per lar complications such as stroke occur early after the 1000 patient-years in those with known CVD versus diagnosis of T2DM [25]. Here, during the five years 5.5 per 1000 patient-years in diabetics with no previous following the initiation of oral hypoglycaemic ther- CVD. apy, 9.1% of 12 272 diabetic patients had a stroke, In general, such epidemiological studies classify representing a twofold risk compared to the general patients as diabetic if the diagnosis were known to their population (Figure 6.1). health care provider prior to the incident event. Given The intensive treatment of T2DM can be shown to that the prevalence of unrecognized T2DM in the older result in an improved maintenance of euglycaemia, but population may be as high as 14.8%, the impact of DM the clinical benefits of such a strategy are largely con- on stroke incidence is likely to be underestimated. fined to the prevention of microvascular complications. The UKPDS was a randomized controlled trial that examined the effect of intensive blood glucose control 6.2.1 Control of DM and stroke risk with either sulphonylureas or insulin compared with conventional care upon the risk of both microvascular Epidemiological studies provide conflicting evidence and macrovascular complications [26]. Intensive treat- regarding the association between chronic hypergly- ment resulted in a significant 11% reduction in HbA1C caemia and stroke risk [22]. However, there appears that was maintained throughout 10 years of follow-up. to be a relationship between hyperglycaemia as mea- However, this improvement in glucose control was not suredbyHbA1C and incident stroke, with a threshold associated with a reduction in the rate of stroke, with ≥7% [23]. Most interventional studies have shown 5.6% of patients in the intensive treatment group and that improved glucose control in both T1DM and 5.2% in the conventional treatment group having a fa- T2DM can delay only the development and progres- tal or non-fatal stroke. In the intensive treatment group sion of microvascular complications such as retinopa- there was a non-significant reduction in myocardial thy, nephropathy and neuropathy. It is, therefore, im- infarction (MI) (Figure 6.2). In addition, within the portant to emphasize that patients with T2DM are intensive treatment group there was a non-significant more likely to develop macrovascular complications of trend (p = 0.07) towards an increase in stroke risk

7000

6000

5000

4000 General Population 3000 DM Cohort

2000

Stroke rate/100000 PY 1000

0 0–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 65–69 70–74 75–79 80–84 85–89 90+ Age (Years)

Figure 6.1 Age-standardized stroke hospitalization rates for the cohort with newly treated type 2 diabetes mellitus (DM) and the 1996 general population of Saskatchewan. PY = patient-years. Reprinted with permission from Ref. [25]. 78 CH 06 STROKE AND DIABETES

30 30 Microvascular endpoints Microvascular endpoints P = 0.099 P = 0.20 between intensive groups

20 20

10 10 Patients with events (%) 0 0

30 30 Myocardial infarction Myocardial infarction P = 0.052 P = 0.66 between intensive groups 20 20

10 10 Patients with events (%) 0 0

30 30 Stroke Stroke P = 0.52 P = 0.07 between intensive groups

20 20

10 10 Patients with events (%) 0 0 036912150 3 6 91215 Time from randomization (years) Time from randomization (years)

Figure 6.2 Kaplan-Meier plots of aggregate end points: microvascular disease, myocardial infarction and stroke for intensive and conventional treatment and by individual intensive therapy in UKPDS 33. Reprinted with permission from Ref. [26].

with glibenclamide (7.3%) compared to either chlor- increasing the risk of hypoglycaemia [27]. In fact, dur- propamide (5.3%) or insulin (4.6%) therapy. ing the long-term follow-up of overweight participants The question is, therefore, as the intensive treatment in UKPDS, who were randomized to diet alone versus of T2DM with either insulin or a sulphonylurea does intensive treatment with metformin, insulin or sulpho- not reduce the risk of cerebrovascular disease, is there a nylurea, additional clinical benefits were observed with metformin therapy [28]. When compared with dietary role for monotherapy with metformin? Evidence from management alone, metformin treatment decreased not the UKPDS indicates that whilst treatment with insulin, only the risk of all-cause mortality by 36%, but also the chlorpropamide or glibenclamide achieves similar de- combined risk of MI, sudden death, angina, peripheral grees of diabetic control, there is the attendant risk vascular disease and stroke by 30%. Furthermore, in of weight gain. In patients with T2DM who are over- comparison with sulphonylurea or insulin, metformin weight, the use of metformin can achieve equivalent was associated with significantly greater reductions levels of glycaemic control, without affecting weight or in any diabetes-related end point, including all-cause 6.3 DIABETES, POST-STROKE HYPERGLYCAEMIA AND PROGNOSIS AFTER ACUTE STROKE 79 mortality and stroke. Although the results of UKPDS with a 2.47-fold increased risk of stroke compared to 34 suggested that the addition of metformin to sulpho- a 3.28-fold increase for diabetes plus the metabolic nylurea therapy might actually increase mortality, it is syndrome [35]. Among 14 000 patients with CHD fol- likely that the need for such combined treatment iden- lowed prospectively for 4.8 to 8.1 years, those with tified those patients whose prognosis was already poor the metabolic syndrome had a 1.49 increased odds for due to factors such as age and poor glycaemic control. ischaemic stroke or TIA [36]. Insulin resistance is a Since the publication of the UKPDS results, ad- precursor to impaired glucose tolerance, and is seen ditional oral antidiabetic drugs have been introduced in both T2DM and the metabolic syndrome. Most epi- that are used as either monotherapy or in combina- demiological studies describe a significant association tion with sulphonylureas or metformin. These include between insulin resistance and stroke risk [22]. The re- meglitinides, alpha-glucosidase inhibitors and thiazo- cent British Women’s Heart and Health Study included lidinediones. The PROspective pioglitAzone Clinical a prospective cohort of 3246 women aged between 60 Trial In macroVascular Events (PROactive) was a ran- and 79 years, who were free of diabetes, stroke or domized controlled trial involving 5238 patients with CHD at baseline. Fasting insulin and a homeostasis T2DM who received the thiazolidinedione drug piogli- model assessment for insulin sensitivity were associ- tazone, or placebo, in addition to their usual cardiovas- ated with stroke and CHD, while fasting blood glucose cular and diabetic medications [29]. A pre-specified and HbA1C were not, indicating a role for insulin resis- subgroup analysis was undertaken of all participants tance as a vascular risk factor [37]. The link between in PROactive with a previous history of stroke; this T2DM, insulin resistance and the metabolic syndrome involved 486 subjects in the treatment arm and 498 must be taken into account in any strategies for vascu- in the placebo group. The use of pioglitazone was lar risk reduction in older diabetics. Specific therapies associated with a trend towards benefit for the pri- for the prevention of cerebrovascular disease in T2DM mary end point of all-cause mortality, major vascular will be considered later in the chapter. events or intervention (p = 0.0670). In this subgroup analysis, active treatment also reduced the rate of fatal and non-fatal stroke (5.6% versus placebo 10.2%, 6.3 Diabetes, post-stroke p = 0.0085). There was some early evidence that the hyperglycaemia and prognosis use of another drug in the same class, rosiglitazone, may be associated with a reduction in the progression after acute stroke of carotid intima-media thickness in T2DM [30]. How- ever, further trials are required to establish the role of Beyond increasing the risk of first or recurrent stroke, this class of drugs in the prevention of cerebrovascu- diabetes is also associated with a poor prognosis fol- lar events in T2DM, particularly given recent concerns lowing the acute event, increasing both mortality and about the safety of thiazolidinediones in patients with dependency [38, 39]. This association also extends to cardiac disease [31]. With respect to the other new oral the finding of hyperglycaemia in the immediate af- hypoglycaemic agents, in the STOP-NIDDM trial us- termath of stroke – the so-called post-stroke hypergly- ing the alpha-glucosidase inhibitor acarbose in patients caemia (PSH) [40–42]. PSH may reflect the metabolic with impaired glucose tolerance, there were very low stress of the acute event, and/or impaired glucose stroke rates in both active and placebo-treated groups metabolism (impaired glucose tolerance or diabetes) [32]. At present, there is no evidence available regard- previously manifest or otherwise. In one series of ing the effect of meglitinide analogues on the long-term UK stroke patients, it was estimated that up to 68% outcomes of T2DM, including macrovascular compli- had PSH, defined by a plasma glucose concentration − cations such as stroke [33]. >6.0 mmol l 1 [43]. The connection between PSH and Individuals with T2DM have a higher prevalence of poor outcome after acute stroke has now been identi- hypertension, central obesity and atherogenic dyslipi- fied in a number of clinical studies, and overall it is daemia [34]. These factors, when clustered together, estimated to be associated with a two- to sixfold in- are termed the metabolic syndrome or syndrome X. creased risk of mortality. In contrast to early reports Whilst diabetes alone increases the risk of stroke, the that PSH and a poor prognosis was confined to patients presence of T2DM as a component of the metabolic with the most severe strokes, other studies have now syndrome leads to a further rise in stroke risk. In the demonstrated that this risk extends across all clinical Framingham Offspring Study, diabetes was associated subtypes of stroke [44]. 80 CH 06 STROKE AND DIABETES

Table 6.1 Admission glycaemic status and stress hyperglycaemia in acute stroke.∗ Admission plasma Admission glycated glucose haemoglobin (HbA1c) Stress hyperglycaemia ↑→ Impaired glucose tolerance/diabetes with ↑↑ stress hyperglycaemia Impaired glucose tolerance/diabetes →↑ without stress hyperglycaemia Normal glucose tolerance without stress →→ hyperglycaemia

Within normal laboratory range → Above normal laboratory range ↑ ∗Adapted from Ref. [40].

PSH may occur due to a number of different mech- association between hyperglycaemia and stroke out- anisms, which may influence – either independently come does not prove cause and effect, the totality of the or in combination – the subsequent clinical outcome. clinical evidence tends to support a direct relationship. PSH was previously thought to be largely due to an acute stress response occurring in association with elevations in plasma cortisol, glucagon, catecholamines 6.4 Diabetes, hyperglycaemia and and often also leukocytosis [45]. Although these acute acute stroke treatment physiological responses may be contributory, more recent clinical data suggest that much of this response is There is still no safe, simple and effective medical ther- associated with an impaired glucose metabolism, with apy that can be given to the majority of acute stroke the prevalence of previously unrecognized diabetes or patients. Even in the absence of a simple medical treat- impaired glucose tolerance preceding stroke being as ment, the benefits of organized and coordinated stroke high as 42%. On the basis of an acute stroke patient’s care have been recognized in terms of reduced mortal- admission plasma glucose and HbA1C level, it is pos- ity, dependency and institutionalization, leading to the sible to determine whether or not stress is a major widespread introduction of specialist stroke services component of their hyperglycaemia (Table 6.1). [47, 48]. Thrombolysis with alteplase (recombinant tis- In patients with PSH and no previous history of dys- sue plasminogen activator, rt-PA) is the only approved glycaemia, the diagnosis of DM or impaired glucose medical therapy for patients with ischaemic stroke. In- tolerance should ideally be made after the stress of the travenous thrombolysis of acute ischaemic stroke with acute ictus has dissipated. In one cohort study, oral glu- alteplase within 3 h of symptom onset has been shown cose tolerance tests (GTTs) were performed 3 months in randomized controlled trials to improve functional after acute stroke in patients presenting with PSH (ad- outcome [49]. More recent data from the European mission plasma glucose >6.0 mmol l−1) [46]. Diabetes SITS-MOST study of the use of alteplase in routine or impaired glucose tolerance were found in 58% of clinical practice confirms similar outcomes to the clin- subjects, with an estimated prevalence of previously ical trials: 11.3% mortality and 54.8% independent at unrecognized diabetes of 16–24%. The finding of an 3 months, as estimated by a modified Rankin score −1 admission plasma glucose >6.0 mmol l plus HbA1C of 0–2. Concerns remain, however, regarding the risks >6.2% was highly predictive of diabetes at 3 months and benefits of alteplase, with a rate of symptomatic after stroke (positive predictive value 80%, negative intracerebral haemorrhage at 7 days of 7.3% [50]. Fur- predictive value 96%). thermore, the routine use of alteplase in the UK at While associations between PSH and clinical present is minimal, with <1% of stroke patients receiv- outcome are seen both in the presence and absence ing this treatment in 2006 [51]. Even in experienced of impaired glucose tolerance or diabetes, the combi- North American stroke centres, less than 20% of po- nation of stress hyperglycaemia and impaired glucose tentially eligible patients receive such treatment [52]. metabolism is likely to be associated with the worst The target of the recent Department of Health Stroke prognosis [40]. However, while demonstrating an Strategy is for 10% of acute stroke patients to receive 6.4 DIABETES, HYPERGLYCAEMIA AND ACUTE STROKE TREATMENT 81 treatment with rt-PA [14]. The problem with throm- in the surgical intensive care patients [59]. In the bolytic therapy for acute stroke is the identification of first DIGAMI study, patients presenting with acute those patients most likely to benefit, and those in whom MI and admission plasma glucose >11.0 mmol l−1 an increased risk may be conferred. Factors such as in- (with or without a previous history of diabetes) were creasing age, stroke severity, raised blood pressure and randomized to an insulin infusion for >24 h, followed extensive cerebral infarction on baseline computed to- by subcutaneous insulin four times daily for more than mography (CT) have been shown to increase the risk 3 months [60]. Such treatment conferred a significant of complications. There is also evidence that hyper- 52% relative reduction in mortality up to 12 months glycaemia (>11.2 mmol l−1) may increase the risk of after the acute event. In view of the uncertainty intracerebral haemorrhage fivefold following routine regarding the relative contributions of the acute GKI thrombolysis. In addition, clinical outcome following infusion or subsequent insulin-based metabolic control thrombolysis is worse in hyperglycaemic patients, even to the overall outcome, the DIGAMI 2 study was after the successful restoration of blood flow [53, 54]. undertaken. This trial randomized 1253 patients with There is some evidence that hyperglycaemia at the time T2DM and acute MI to one of three treatment arms: of acute stroke may predict non-recanalization follow- 24 h GKI followed by long-term subcutaneous insulin; ing thrombolysis. In one study, the finding of a blood 24 h GKI then conventional glucose control; or routine glucose concentration >8.8 mmol l−1 at more than 2 h care [61]. Although DIGAMI 2 failed to replicate the after thrombolysis was an independent predictor of fail- results of the earlier trial, with no benefit being seen ure to recanalize (OR 7.3, 95% CI: 1.3–42.3) [55]. in either of the GKI treatment limbs, the trial included In the absence of a simple early medical therapy ap- only those patients with T2DM in whom the baseline plicable to the majority of stroke patients, an essential glucose was much lower than the original DIGAMI part of acute stroke unit care is the intensive monitoring cohort, and the degree of glucose lowering achieved of physiological variables (hydration, blood glucose, was less. This may partially explain the differing temperature, blood pressure, oxygen saturation). The results obtained from the two studies. RCP National Clinical Guidelines for Stroke empha- Such a prolonged insulin treatment regimen is not size the need to consider the early management of feasible in the majority of acute stroke units, where the hyperglycaemia, hypertension, hydration and pyrexia complexity of clinical care combined with the practical [12]. Whilst there is accumulating evidence for a link difficulties in maintaining hydration and nutrition may between hyperglycaemia, diabetes and enhanced cere- make routine treatment with insulin beyond 24 h un- bral ischaemic damage, until recently intervention to safe. Furthermore, whilst hyperglycaemia may be seen modulate hyperglycaemia in acute stroke had not been in the majority of acute stroke patients, this is usu- examined in a randomized controlled trial. ally mild with mean plasma glucose concentrations of Early clinical trials in acute MI and critically ill 8–9 mmol l−1 [62]. Glucose control following stroke is patients in the intensive care unit support the concept further complicated by the fact that up to one-third of of treating hyperglycaemia and maintaining eugly- patients have dysphagia in the immediate aftermath of caemia with insulin. A variety of methods for insulin the acute event. In two-thirds of cases, these swallow- administration is available in these settings, and both ing difficulties usually resolve within the first week; sliding-scale insulin and glucose-potassium-insulin however, nutritional support and supplementation may (GKI) regimes have potential disadvantages [56]. be fraught with practical and ethical problems. There An early overview of trials in acute MI has shown is accumulating evidence that PSH is maximal in the that treatment with a GKI-based regimen reduces initial 12–18 h after stroke, and that glucose levels in-patient mortality by 28% [57]. In addition, the will decline spontaneously without specific interven- use of insulin after acute MI may confer survival tion [62]. benefits, even in the absence of initial hyperglycaemia Treatment with a variable-dose GKI infusion can [58]. Consistent with these findings is evidence from safely induce and maintain euglycaemia during the a recent meta-analysis of 35 randomized controlled first 24 h in hospital. The United Kingdom Glucose trials concerning the effects of insulin in 8478 Insulin in Stroke Trial (GIST-UK) was a pragmatic critically ill patients with hyperglycaemia. In the multicentre, randomized controlled trial that sought to majority of these studies, insulin was administered determine whether outcome from acute stroke could as a GKI infusion. Treatment with insulin decreased be favourably influenced by GKI-induced and main- mortality by 15%, with the greatest benefit seen tained euglycaemia when delivered as part of routine 82 CH 06 STROKE AND DIABETES stroke unit care [63]. Patients presenting within 24 h should be remembered, however, that the glucose in- of symptom onset of ischaemic stroke or primary intervention in GIST-UK was short-term, whereas in tracerebral haemorrhage were randomized to receive DIGAMI the treatment was continued for a mini- either variable-dose GKI or saline as a continuous in- mum of 12 weeks [60]. Furthermore, GIST-UK was travenous infusion for 24 h. The purpose of the GKI a pragmatic trial designed to be carried out in NHS treatment was to maintain capillary glucose levels at acute stroke units, where the much more intensive glu- 4–7 mmol l−1, with no glucose-lowering intervention cose intervention and monitoring seen in ICCU trials in the control group. Patients with insulin-treated DM could not be achieved [64]. One criticism of GIST-UK were excluded from the trial, and participants had when compared to DIGAMI was that the intensity of only modest degrees of PSH (median baseline glucose glucose-lowering was insufficient to influence the out- 7.6 mmol l−1, IQR 6.7–9.0). The primary outcome for come. A post hoc analysis of the GIST population the trial was mortality at 90 days, with a secondary found that intensive lowering of glucose (>2 mmol outcome of avoidance of death or severe disability. l−1) between baseline and 24 h with GKI was asso- Recruitment into the GIST-UK study was halted due ciated with a 34% excess mortality when compared to slow enrolment, at which time 933 patients had been to patients with a glucose reduction <2 mmol l−1.Al- included. There was no significant reduction in mor- though this finding was reported with caution, recent tality at 90 days (GKI versus control OR 1.14, 95% evidence obtained with positron emission tomography CI: 0.86–1.51, p = 0.37), nor any significant differ- (PET) scanning suggested that hyperglycaemia might ences for secondary functional outcomes, as measured not be directly harmful to the ischaemic brain, and in using the modified Rankin scale and Barthel index the presence of cerebral ischaemia lactate derived from (Figure 6.3). In the GKI treatment group, both overall anaerobic metabolism may be the preferred energy sup- mean plasma glucose and mean systolic blood pressure ply [65]. Thus, an intensive reduction of glucose may were significantly lower than in the control arm (mean in fact reduce glucose load to the brain and attenuate difference in glucose 0.57 mmol l−1,p< 0.001; mean lactate production. difference in blood pressure 9.0 mmHg, p < 0.0001) As the first clinical trial of glucose modulation (Figure 6.4). The hypotensive effect of GKI treat- in acute stroke, GIST-UK was planned during the ment was unexpected, and might in part have been 1990s, and it is encouraging that other clinical trials due to the potassium component of the infusion or, are currently being undertaken in this area (GRASP conversely, the result of a relative pressor effect of trial, www.grasptrial.org; IRIS trial, www.iristrial.org; intravenous saline. The results of GIST-UK were at NICE and SUGAR study) Alternative methods of in- variance with the results of trials of glucose-lowering sulin administration were examined in the recently in coronary care and intensive care unit settings. It published Treatment of Hyperglycaemia in Ischaemic

GKI Saline n/total n/total

Barthel Index <9 73/309 88/327 0.84 (0.59−1.20)

mRS >3 (excluding death) 113/322 122/339 0.96 (0.70−1.32)

Death 139/464 128/469 1.14 (0.86−.51)

0.4 0.6 0.8 1.0 1.2 1.4 1.6 Favours GKI Favours saline

Figure 6.3 Glucose insulin in stroke trial (GIST-UK). Common odds ratios (ORs) for primary and secondary outcomes with 95% CIs in the intention-to-treat data set. GKI = glucose-potassium-insulin. Reprinted with permission from Ref. [63]. 6.4 DIABETES, HYPERGLYCAEMIA AND ACUTE STROKE TREATMENT 83

8.5 GKI Saline

8.0

7.5

7.0 Glucose (mmol/L)

6.5

6.0

165 GKI Saline

160

155

Systolic Blood Pressure (mm Hg) 150

145 Admission 8 hours 16 hours 24 hours Time since start of treatment

Figure 6.4 Glucose insulin in stroke trial (GIST-UK). Mean plasma glucose concentration and systolic blood pressure at baseline, 8 h, 16 h and 24 h. Reprinted with permission from Ref. [63].

Stroke (THIS) trial [66]. In this pilot study, 46 patients seen to be significantly lower in the aggressive treat- were randomized within 12 h of cerebral infarction to ment group (7.4 versus 10.5 mmol l−1,p< 0.001), receive either aggressive treatment with continuous in- with hypoglycaemia occurring in 35% of these patients. travenous insulin (target glucose <7.2 mmol l−1) or Unfortunately, the small number of patients included in usual treatment with subcutaneous insulin four times this pilot study precluded any assessment of the clinical daily (target <11.1 mmol l−1). The glucose levels were efficacy of this treatment strategy. 84 CH 06 STROKE AND DIABETES Clinical guidelines emphasize the role of physiolog- both diabetic and non-diabetic individuals. The results ical monitoring and intervention in acute stroke care of epidemiological studies have shown that usual sys- [12]. However, thresholds for the routine treatment tolic and diastolic blood pressure levels are directly of hyperglycaemia vary across centres and between and continuously associated with a risk of both cere- clinicians, with varying local policies for the man- bral infarction and primary intracerebral haemorrhage agement of PSH. The physiological effect of interven- in patients with and without a history of hyperten- tions such as saline hydration or glucose modulation is sion [68, 69]. Furthermore, a similar linear relationship poorly understood, and requires further investigation. exists between systolic and diastolic blood pressure In GIST-UK, a simple saline infusion, which is part of and the risk of recurrent cerebrovascular events fol- routine stroke unit care, was associated with a signifi- lowing stroke or TIA [70]. Reducing diastolic blood cant reduction in plasma glucose. However, the small pressure by 5–6 mmHg in people with hypertension reduction in glucose concentration achieved with GKI and no history of cerebrovascular disease decreases was not associated with a net clinical improvement, their risk of stroke by approximately one-third, with and any benefit from such reductions may be offset by all major classes of antihypertensive agents appearing a greater decrease in blood pressure. The challenge for equally effective [71]. In addition, pharmacological in- future trials in acute stroke will be to determine the tervention to lower blood pressure has been shown to safety and efficacy of glucose-lowering in an appropri- reduce the risk of stroke recurrence in hypertensive ate high-dependency environment, while maintaining stroke survivors [72]. Although hypertension is a ma- other physiological variables such as blood pressure jor determinant of stroke risk in diabetics, most patients within a satisfactory range. Despite the absence of con- have other risk factors such as dyslipidaemia and is- clusive clinical trial evidence, in patients for whom chaemic heart disease that may influence the choice of thrombolysis is contemplated, it would seem prudent antihypertensive therapy. to attempt the correction of hyperglycaemia, although Embedded within the United Kingdom Prospective the therapeutic time window for thrombolytic therapy Diabetes Study was a randomized controlled trial is so short that effective glycaemic control may actually (UKPDS 38) to establish if the tight control of follow rather than precede such treatment. blood pressure (<150/85 mmHg) decreased morbidity and mortality in patients with T2DM [73]. Treated 6.5 Stroke prevention in type 2 hypertensive diabetics whose blood pressure was above this target level and those who were previously diabetes untreated (≥160/≥90 mmHg) received either intensive or less-intensive blood pressure-lowering therapy. Diabetes and hyperglycaemia are not only major risk The intensive treatment comprised an angiotensin- factors for stroke but also important prognostic fac- converting enzyme (ACE) inhibitor (captopril) or a tors for clinical outcome. After the initial phase of the β-blocker (atenolol), with a target blood pressure of illness, the intensive management of vascular risk fac- <150/85 mmHg. The purpose of the other treatment tors confers major benefits for stroke patients. Until a limb was a less tight control of blood pressure safe, simple and effective acute therapy is developed (<180/105 mmHg), avoiding ACE inhibitors and for the majority of stroke patients, however, the priority β-blockers. Almost one-third (29%) of patients must be to reduce mortality, disability and dependency randomized to the tight control group needed three or through the implementation of proven primary and sec- more agents to control their blood pressure, compared ondary preventive strategies. to only 11% in the less-intensive treatment group. A The latest guidelines on the management of CVD tight control resulted in a significantly lower blood in diabetes emphasize that stroke prevention should be pressure; the mean blood pressure over 9 years of based on a multifactorial strategy, including the treat- follow-up was 144/82 mmHg in the intensive treatment ment of hypertension, hyperglycaemia, microalbumin- group compared to 154/87 mmHg in the other limb. uria, hyperlipidaemia, and also the use of antiplatelet After a median follow-up period of 8.4 years, the in- medication [67]. tensive management of hypertension in these UKPDS patients resulted in a 24% relative risk reduction for the 6.5.1 Hypertension development of any diabetes-related end point. There Hypertension is the single most important potentially was no significant difference in diabetic control be- reversible risk factor for cerebrovascular disease in tween the two blood pressure treatment groups (mean 6.5 STROKE PREVENTION IN TYPE 2 DIABETES 85

HbA1C 7.2% for both). Intensive treatment resulted in protective effect on the vasculature of drugs such as a significant 44% relative reduction in fatal or non-fatal ACE inhibitors, or an enhanced control of ambulatory stroke, but no significant decrease in MI. These results blood pressure levels compared with clinic readings were similar to those seen with other trials of blood [83]. pressure-lowering treatment in older people [74, 75]. Further evidence for the beneficial effect of ACE in- Thus, intensive glycaemic control alone in T2DM is hibitors in reducing the risk of recurrent stroke comes not sufficient to reduce risk of stroke; simultaneous from the PROGRESS trial [84]. This study recruited management of hypertension is also necessary. 6105 hypertensive and normotensive patients with a UKPDS 38 also showed that intensive blood pres- history of cerebrovascular disease, of whom 13% were sure treatment necessitated combination therapy. It diabetic. The maximum beneficial effect was seen should be noted that even in the UKPDS 38 tight con- in patients on combined therapy with the ACE in- trol group, the mean level of blood pressure achieved hibitor perindopril plus the diuretic, indapamide. Mean was still higher than the current UK recommended blood pressure lowering on combined therapy was target of <130/80 mmHg for clinic readings [76]. How- 12/5 mmHg compared with perindopril monotherapy ever, this trial did demonstrate that the lowest risk of (mean 5/3 mmHg). Treatment with perindopril plus in- complications due to diabetes was seen in patients with dapamide resulted in a relative risk reduction for re- a systolic blood pressure <120 mmHg, in line with current stroke of 43%, compared to a non-significant current recommendations [77]. The choice of antihy- 5% decrease with perindopril alone. However, no sub- pertensive therapy for patients with T2DM will also group analysis was performed for the diabetic pa- be influenced by the results of more recent clinical tients enrolled in this study. Additional data to sup- trials. In the ASCOT study, 19 257 patients with hy- port the use of ACE inhibitors in T2DM derive from pertension and three other cardiovascular risk factors the ADVANCE trial [85], where 11 140 patients with were randomized to receive amlodipine (plus perindo- T2DM were randomized to receive perindopril plus pril if required) or atenolol (plus a thiazide diuretic) indapamide along with their usual treatment, irrespec- [78]. The amlodipine-based regime prevented more tive of their initial blood pressure. A modest reduc- cardiovascular events and induced less diabetes than tion in blood pressure of 5.6/2.2 mmHg was seen with the atenolol-based regime. In the ASCOT study, it is perindopril/indapamide, but there was a 9% reduction possible that differing effects of the two treatment arms in risk of all major macrovascular or microvascular on variables other than blood pressure contributed to events as well as a decrease in all-cause mortality. the different event rates, particularly for stroke [79]. In hypertensive diabetic patients, the intensity of There is accumulating evidence from ASCOT and blood pressure therapy is important in order to maxi- other studies that, for patients at high risk of vascular mally decrease their risk of stroke and other vascular disease, the benefits of antihypertensive therapy may events. In order to further investigate the intensity of extend beyond blood pressure lowering. In the HOPE blood pressure lowering, the HOT study examined the study, 9297 high-risk vascular patients aged ≥55 years, optimum treatment level for diastolic blood pressure. including 3577 (38%) with diabetes plus one additional The study found that there was a twofold increase in risk factor (including stroke or TIA), received treat- the incidence of cardiovascular events in hypertensive ment with the ACE inhibitor ramipril or placebo [80]. diabetic patients, and that in these individuals an inten- Ramipril resulted in modest reductions in office blood sive treatment to reduce the diastolic blood pressure to pressure compared to placebo (3.8/2.8 mmHg). Never- <80 mmHg resulted in a 30% reduction in the risk theless, the overall results confirmed that the relative of stroke compared with more modest reductions in risk of any stroke or fatal stroke was decreased by blood pressure to <90 mmHg [86]. Treatment in the 32% and 61%, respectively. Ramipril was beneficial HOT study was with a calcium channel blocker-based even in those who were normotensive at baseline. A regime, but for many patients this was combined with further analysis of the results for the 3577 diabetic par- a β-blocker or ACE inhibitor. ticipants showed a significant 33% reduction in stroke Intensive blood pressure-lowering in T2DM is be- risk, again irrespective of baseline blood pressure [81]. ing evaluated as part of the on-going Action to Con- Thus, as in the UKPDS, the benefits observed with trol Cardiovascular Risk in Diabetes (ACCORD) trial fairly modest reductions in office blood pressure were [87]. The blood pressure limb of ACCORD is com- greater than might be predicted from epidemiological paring target systolic pressures of <120 mmHg versus data [82]. Explanations for this observation include a <140 mmHg. 86 CH 06 STROKE AND DIABETES Current international management guidelines recom- known ischaemic heart disease, cholesterol-lowering mend a target systolic blood pressure of <130 mmHg therapy was associated with a 30% reduction in the for diabetic patients [67]. However, accumulating ev- risk of stroke [92]. The beneficial effects of statins idence infers that high-risk vascular patients such as were observed in people with only moderately raised diabetics should be treated even more aggressively, ir- cholesterol levels. Furthermore, statins did not seem to respective of baseline blood pressure, in order to reduce increase the risk of haemorrhagic stroke, despite the the risk of cerebrovascular events. The management of epidemiological evidence for an association between hypertension immediately after acute stroke is more low cholesterol and intracerebral haemorrhage. contentious [88]. Most acute stroke patients are hyper- The Cholesterol Treatment Trialists’ Collaboration tensive on admission to hospital, and in many cases has recently published a meta-analysis focusing on the the blood pressure falls spontaneously over the first use of statins in diabetic patients [93]. 7–10 days. The cerebral autoregulation of blood flow Data from 14 randomized controlled trials were used is impaired early after stroke, and cerebral perfusion in this meta-analysis, which included 18 686 diabet- may become dependent upon systemic blood pressure ics, 17 220 of whom had T2DM. The analysis was levels. Thus, lowering the blood pressure in the acute undertaken in the context of the 71 370 non-diabetic phase after stroke could be potentially harmful, and participants in the same studies. Over a mean follow-up on-going clinical trials are addressing this issue. In period of 4.3 years, a total of 3247 vascular events the PROGRESS and HOPE trials, ACE inhibitor ther- occurred in the diabetic patients. A 9% proportional apy was delayed for 2–4 weeks after stroke. Based reduction in all-cause mortality per mmol l−1 decrease on current evidence, it would seem prudent to within LDL-cholesterol was observed, similar to the 13% hold antihypertensive therapy until stroke patients are reduction seen in non-diabetic individuals. Among di- clinically and neurologically stable, except for specific abetics, there was a significant decrease in vascular clinical circumstances such as accelerated hypertension deaths (rate ratio 0.87, p = 0.008), with no effect where immediate blood pressure-lowering is manda- seen on non-vascular mortality. Furthermore, in peo- tory [89]. ple with diabetes there was a significant decrease in stroke (rate ratio 0.79, p = 0.0002), with significant 6.5.2 Dyslipidaemia reductions also seen for MI and coronary death or revascularisation (Figure 6.5). Among diabetics, the The relationship between dyslipidaemia and CVD is proportional effects of statins were similar, irrespec- qualitatively similar in diabetic and non-diabetic pa- tive of any past history of vascular disease or other tients, but for any given level of cholesterol the ab- baseline characteristics. These findings strengthen the solute risk is higher for diabetics. The evidence that evidence for the use of statin therapy in all diabetic dyslipidaemia is a risk factor for stroke is conflict- patients at high risk of vascular disease. The lipid limb ing, with a recent meta-analysis showing no apparent of the on-going ACCORD trial includes 5518 partici- association [68]. Many studies, however, do not dis- pants with T2DM who are receiving either fenofibrate tinguish between cerebral infarction and primary in- or placebo to determine whether the use of a fibrate tracerebral haemorrhage. Evidence from Caucasian and to increase HDL-cholesterol and lower triglycerides, Asian population studies shows a positive association together with a statin to lower LDL-cholesterol, will between total cholesterol levels and the risk of cere- reduce the risk of vascular events [87]. bral infarction, possibly offset by a negative correlation between cholesterol and risk of intracerebral haemor- 6.5.3 Antiplatelet therapy rhage [90, 91]. It should be remembered that the majority of diabetic Antiplatelet therapy is an important part of secondary patients at risk of stroke are older and have T2DM. In prevention in patients with TIA or stroke. The these individuals, the predominant lipid abnormalities Antiplatelet Trialists’ Collaboration meta-analysis are raised triglycerides and reduced HDL-cholesterol. confirmed a significant reduction in vascular events In contrast to the evidence linking dyslipidaemia to (non-fatal MI, non-fatal stroke, vascular death) in ischaemic heart disease, it is only more recently that diabetic patients with vascular disease who were statin therapy has also been shown to be effective in treated with antiplatelet agents [94]. The results of reducing cerebrovascular disease. A meta-analysis of the recent ESPRIT trial, in combination with those the early statin trials confirmed that, in patients with of previous trials, provide evidence for the use of 6.6 CONCLUSIONS 87

Major vascular event Events (%) and prior diabetes Treatment Control RR(Cl)

Major coronary event Diabetes 776(8.3%) 979(10.5%) 0.78(0.69−0.87) No diabetes 2561(7.2%) 3441(9.6%) 0.77(0.73−0.81) Any major coronary event 3337(7.4%) 4420(9.8%) 0.77(0.74−0.80) χ2 Test for heterogeneity within subgroup: 1 = 0.1; p = 0.8

Coronary revascularisation Diabetes 491(5.2%) 627(6.7%) 0.75(0.64−0.88) No diabetes 2129(6.0%) 2807(7.9%) 0.76(0.72−0.81) Any coronary revascularisation 2620(5.8%) 3434(7.6%) 0.76(0.73−0.80) χ2 Test for heterogeneity within subgroup: 1 = 0.1; p = 0.8

Stroke Diabetes 407(4.4%) 501(5.4%) 0.79(0.67−0.93) No diabetes 933(2.7%) 1116(3.2%) 0.84(0.76−0.93) Any stroke 1340(3.0%) 1617(3.7%) 0.83(0.77−0.88) χ2 Test for heterogeneity within subgroup: 1 = 0.8; p = 0.4

Major vascular event Diabetes 1465(15.6%) 1782(19.2%) 0.79(0.72−0.86) No diabetes 4889(13.7%) 6212(17.4%) 0.79(0.76−0.82) Any major vascular event 6354(14.1%) 7994(17.8%) 0.79(0.77−0.81) χ2 Test for heterogeneity within subgroup: 1 = 0.0; p = 0.9

RR (99% Cl) 0.5 1.0 1.5 RR (95% Cl) Treatment better Control better

Figure 6.5 Cholesterol treatment trialists’ collaboration. Proportional effects on major vascular events per mmol l−1 reduction in LDL-cholesterol in participants with or without diabetes. RR = relative risk. Reprinted with permission from Ref. [93]. aspirin plus dipyridamole after ischaemic stroke in MATCH study population who were diabetic, aspirin preference to aspirin alone [95]. Current clinical plus clopidogrel produced a slightly higher relative guidelines therefore recommend the use of aspirin risk reduction than in the remainder of the study plus modified-release dipyridamole for the secondary population, but this again was not significant and was prevention of occlusive vascular events following associated with increased bleeding. ischaemic stroke or TIA [12, 96]. The MATCH trial assessed the role of combined antiplatelet therapy with aspirin plus clopidogrel in patients with ischaemic 6.6 Conclusions cerebrovascular disease, with a combined vascular end point of ischaemic stroke and rehospitalization for Older patients with T2DM are at high risk of macrovas- TIA [97]. The combination of aspirin plus clopidogrel cular complications, including stroke and TIA. In ad- produced a marginally increased relative risk reduction dition, cerebrovascular disease is a major contributor when compared to aspirin, which did not achieve to cognitive decline in this population. A reduction in statistical significance and was offset by a higher risk cerebrovascular disease can be achieved by a com- of bleeding in this population. Among the 68% of the prehensive strategy of vascular risk reduction aimed 88 CH 06 STROKE AND DIABETES at multiple risk factors [98]. On-going clinical trials of ischemic stroke: importance of population-based such as ACCORD should clarify the risks and benefits studies. Stroke, 34, 2050–9. of a strategy of intensive glucose and blood pressure 10. Isarol, P. A. and Forbes, J. F. (1992) The cost of stroke control plus management of lipids with statins and fi- to the National Health Service in Scotland. Cerebrovasc brates; the glucose-lowering limb of this study was Dis, 1, 47–50. recently stopped 18 months early due to an increased 11. Macdonald, B. K., Cockerell, O. C., Sander, J. W. mortality when blood glucose concentrations were low- and Shorvon, S. D. (2000) The incidence and lifetime ered below current recommended levels [87]. Based prevalence of neurological disorders in a prospective on the available evidence, treatment of the older per- community-based study in the UK. Brain, 123 (Pt 4), son with T2DM should encompass lifestyle changes, a 665–76. tight control of glycaemia and hypertension, and tar- 12. Intercollegiate Stroke Working Party. (2008) National geted specific drug therapies such as ACE inhibitors, Clinical Guidelines for Stroke, 3rd edition. London, statins and antiplatelet agents. Royal College of Physicians. 13. Department Of Health (2001) National Service Frame- work for Older People. Department of Health, London. 14. Department Of Health (2007) National Stroke Strategy. References Department of Health, London. 15. Wahl, P. W., Savage, P. J., Psaty, B. M., Orchard, 1. Kannel, W. B. and McGee, D. L. (1979) Diabetes and T. J., Robbins, J. A. and Tracy, R. P. (1998) Dia- cardiovascular disease. The Framingham study. JAMA, betes in older adults: comparison of 1997 American 241, 2035–8. Diabetes Association classification of diabetes mellitus 2.Fuller,J.H.,Shipley,M.J.,Rose,G.,Jarrett,R.J. and Keen, H. (1983) Mortality from coronary heart with 1985 WHO classification. Lancet, 352, 1012–15. disease and stroke in relation to degree of glycaemia: 16. Gunarathne, A., Patel, J. V., Potluri, R., Gammon, the Whitehall study. Br Med J (Clin Res Ed.), 287, B., Jessani, S., Hughes, E. A. and Lip, G. Y. (2008) 867–70. Increased 5-year mortality in the migrant South Asian 3. World Health Organization (1971) Cerebrovascular stroke patients with diabetes mellitus in the United Diseases: Prevention, Treatment and Rehabilitation. Kingdom: The West Birmingham Stroke Project. Int Report of a WHO meeting. World Health Organ Tech J Clin Pract., 62 (2), 197–201. Rep Ser, 469, 1–57. 17. Laakso, M. (1999) Hyperglycaemia and cardiovascular 4. Albers, G. W., Caplan, L. R., Easton, J. D., Fayad, disease in Type 2 diabetes. Diabetes, 48, 937–42. P.B.,Mohr,J.P.,Saver,J.L.andSherman,D.G. 18. Coutinho, M., Gerstein, H. C., Wang, Y. and Yusuf, S. (2002) Transient ischemic attack - proposal for a new (1999) The relationship between glucose and incident definition. N Engl J Med, 347, 1713–6. cardiovascular events. A metaregression analysis of 5. Adams, H. P., Jr., Bendixen, B. H., Kappelle, L. J., published data from 20 studies of 95,783 individuals Biller, J., Love, B. B., Gordon, D. L. and Marsh, E. E., followed for 12.4 years. Diabetes Care, 22, 233–40. 3rd (1993) Classification of subtype of acute ischemic 19. Mulnier, H. E., Seaman, H. E., Raleigh, V. S., stroke. Definitions for use in a multicenter clinical trial. Soedamah-Muthu, S. S., Colhoun, H. M., Lawrenson, TOAST. Trial of Org 10172 in Acute Stroke Treatment. R. A. and De Vries, C. S. (2006) Risk of stroke in Stroke, 24, 35–41. people with type 2 diabetes in the UK: a study using 6. De Angelis, M. (2003) Prevalence of carotid stenosis in the General Practice Research Database. Diabetologia, type 2 diabetic patients asymptomatic for cerebrovas- 49, 2859–65. cular disease. Diabetes Nutrition and Metabolism, 16, 20. Janghorbani, M., Hu, F. B., Willett, W. C., Li, T. Y., 48–55. 7. You, R., McNeil, J. J., O’Malley, H. M., Davis, S. Manson, J. E., Logroscino, G. and Rexrode, K. M. M. and Donnan, G. A. (1995) Risk factors for lacunar (2007) Prospective study of type 1 and type 2 diabetes infarction syndromes. Neurology, 45, 1483–7. and risk of stroke subtypes: the Nurses’ Health Study. 8. Karapanayiotides, T., Piechowski-Jozwiak, B., Van Diabetes Care, 30, 1730–5. Melle, G., Bogousslavsky, J. and Devuyst, G. (2004) 21. Giorda, C. B., Avogaro, A., Maggini, M., Lombardo, Stroke patterns, etiology, and prognosis in patients with F., Mannucci, E., Turco, S., Alegiani, S. S., Raschetti, diabetes mellitus. Neurology, 62, 1558–62. R., Velussi, M. and Ferrannini, E. (2007) Incidence and 9. Schulz, U. G. and Rothwell, P. M. (2003) Differences risk factors for stroke in type 2 diabetic patients: the in vascular risk factors between etiological subtypes DAI study. Stroke, 38, 1154–60. 6.6 CONCLUSIONS 89

22. Air, E. L. and Kissela, B. M. (2007) Diabetes, the 32. Chiasson, J. L., Josse, R. G., Gomis, R., Hanefeld, M., metabolic syndrome, and ischemic stroke: epidemi- Karasik, A. and Laakso, M. (2003) Acarbose treatment ology and possible mechanisms. Diabetes Care, 30, and the risk of cardiovascular disease and hyperten- 3131–40. sion in patients with impaired glucose tolerance: the 23. Myint, P. K., Sinha, S., Wareham, N. J., Bingham, S. STOP-NIDDM trial. JAMA, 290, 486–94. A., Luben, R. N., Welch, A. A. and Khaw, K. T. (2007) 33. Black, C., Donnelly, P., McIntyre, L., Royle, P. L., Glycated hemoglobin and risk of stroke in people Shepherd, J. P. and Thomas, S. (2007) Meglitinide ana- without known diabetes in the European Prospective logues for type 2 diabetes mellitus. Cochrane Database Investigation into Cancer (EPIC)-Norfolk prospective Syst Rev, CD004654. population study: a threshold relationship? Stroke, 38, 34. Gil-Nunez, A. (2007) The metabolic syndrome and 271–5. cerebrovascular disease: suspicion and evidence. Cere- 24. Turner, R., Cull, C. and Holman, R. (1996) United brovasc Dis, 24 (Suppl. 1), 64–75. Kingdom Prospective Diabetes Study 17: a 9-year 35. Najarian, R. M., Sullivan, L. M., Kannel, W. B., Wil- update of a randomized, controlled trial on the ef- son, P. W., D’Agostino, R. B. and Wolf, P. A. (2006) fect of improved metabolic control on complications Metabolic syndrome compared with type 2 diabetes in non-insulin-dependent diabetes mellitus. Ann Intern mellitus as a risk factor for stroke: the Framingham Med, 124, 136–45. Offspring Study. Arch Intern Med, 166, 106–11. 25. Jeerakathil, T., Johnson, J. A., Simpson, S. H. and 36. Koren-Morag, N., Goldbourt, U. and Tanne, D. (2005) Majumdar, S. R. (2007) Short-term risk for stroke Relation between the metabolic syndrome and ischemic is doubled in persons with newly treated type 2 di- stroke or transient ischemic attack: a prospective cohort abetes compared with persons without diabetes: a study in patients with atherosclerotic cardiovascular population-based cohort study. Stroke, 38, 1739–43. disease. Stroke, 36, 1366–71. 26. United Kingdom Prospective Diabetes Study Group. 37. Lawlor, D. A., Fraser, A., Ebrahim, S. and Smith, G. (1998) Intensive blood glucose control with sulphony- D. (2007) Independent associations of fasting insulin, lureas or insulin compared with conventional treatment glucose, and glycated haemoglobin with stroke and and risk of complications in patients with type 2 dia- coronary heart disease in older women. PLoS Med,4, betes (UKPDS 33). Lancet, 352, 837–53. e263. 27. United Kingdom Prospective Diabetes Study Group. 38. Pulsinelli, W. A., Levy, D. E., Sigsbee, B., Scherer, P. (1995) United Kingdom Prospective Diabetes Study and Plum, F. (1983) Increased damage after ischemic (UKPDS) 13: Relative efﬁcacy of randomly allocated stroke in patients with hyperglycemia with or without diet, sulphonylurea, insulin or metformin in patients established diabetes mellitus. Am J Med, 74, 540–4. with newly diagnosed non-insulin dependent diabetes 39. Oppenheimer, S. M., Hoffbrand, B. I., Oswald, G. A. followed for three years. Br Med J, 310, 83–8. and Yudkin, J. S. (1985) Diabetes mellitus and early 28. United Kingdom Prospective Diabetes Study Group. mortality from stroke. Br Med J (Clin Res Ed), 291, (1998) Effect of intensive blood glucose control with 1014–15. metformin on complications in overweight patients 40. Gray, C. S., Taylor, R., French, J. M., Alberti, K. G., with type 2 diabetes. (UKPDS 34). Lancet, 352, Venables, G. S., James, O. F., Shaw, D. A., Cartlidge, 854–65. N. E. and Bates, D. (1987) The prognostic value 29. Wilcox, R., Bousser, M. G., Betteridge, D. J., Schern- of stress hyperglycaemia and previously unrecognized thaner, G., Pirags, V., Kupfer, S. and Dormandy, J. diabetes in acute stroke. Diabet Med, 4, 237–40. (2007) Effects of pioglitazone in patients with type 2 41. Weir, C. J., Murray, G. D., Dyker, A. G. and Lees, K. R. diabetes with or without previous stroke: results from (1997) Is hyperglycaemia an independent predictor of PROactive (PROspective pioglitAzone Clinical Trial In poor outcome after acute stroke? Results of a long-term macroVascular Events 04). Stroke, 38, 865–73. follow up study. Br Med J, 314, 1303–6. 30. Hedblad, B., Zambanini, A., Nilsson, P., Janzon, L. 42. Power, M. J., Fullerton, K. J. and Stout, R. W. (1988) and Berglund, G. (2007) Rosiglitazone and carotid IMT Blood glucose and prognosis of acute stroke. Age progression rate in a mixed cohort of patients with type Ageing, 17, 164–70. 2 diabetes and the insulin resistance syndrome: main 43. Scott, J. F., Robinson, G. M., French, J. M., O’Connell, results from the Rosiglitazone Atherosclerosis Study. J J. E., Alberti, K. G. and Gray, C. S. (1999) Prevalence Intern Med, 261, 293–305. of admission hyperglycaemia across clinical subtypes 31. Macfarlane, D. P. and Fisher, M. (2006) Thiazolidine- of acute stroke. Lancet, 353, 376–7. diones in patients with diabetes mellitus and heart fail- 44. Gray, C. S., O’Connell, J. E. and Lloyd, H. (2001) ure: implications of emerging data. Am J Cardiovasc Diabetes, hyperglycaemia and recovery from stroke. Drugs, 6, 297–304. Geriatrics and Gerontology International,1,2–7. 90 CH 06 STROKE AND DIABETES

45. O’Connell, J. E. and Gray, C. S. (1991) The stress state is associated with lower tPA-induced recanaliza- response to acute stroke. Stress Medicine, 7, 239–43. tion rates in stroke patients. Stroke, 36, 1705–9. 46. Gray, C. S., Scott, J. F., French, J. M., Alberti, K. G. 56. Sawin, C. T. (1997) Action without benefit. The sliding and O’Connell, J. E. (2004) Prevalence and prediction scale of insulin use. Arch Intern Med, 157, 489. of unrecognised diabetes mellitus and impaired glu- 57. Fath-Ordoubadi, F. and Beatt, K. J. (1997) cose tolerance following acute stroke. Age Ageing, 33, Glucose-insulin-potassium therapy for treatment 71–7. of acute myocardial infarction: an overview of 47. Stroke Unit Trialists’ Collaboration (1997) Collabo- randomized placebo-controlled trials. Circulation, 96, rative systematic review of the randomised trials of 1152–6. organised inpatient (stroke unit) care after stroke. Br 58. Diaz, R., Paolasso, E. A., Piegas, L. S., Tajer, C. Med J, 314, 1151–9. D., Moreno, M. G., Corvalan, R., Isea, J. E. and 48. Seenan, P., Long, M. and Langhorne, P. (2007) Stroke Romero, G. (1998) Metabolic modulation of acute my- units in their natural habitat: systematic review of ocardial infarction. The ECLA (Estudios Cardiologicos observational studies. Stroke, 38, 1886–92. Latinoamerica) Collaborative Group. Circulation, 98, 49. Hacke, W., Donnan, G., Fieschi, C., Kaste, M., Von 2227–34. Kummer,R.,Broderick,J.P., Brott, T., Frankel, M., 59. Pittas, A. G., Siegel, R. D. and Lau, J. (2004) In- Grotta, J. C., Haley, E. C., Jr., Kwiatkowski, T., Levine, sulin therapy for critically ill hospitalized patients: a S. R., Lewandowski, C., Lu, M., Lyden, P., Marler, J. meta-analysis of randomized controlled trials. Arch In- R., Patel, S., Tilley, B. C., Albers, G., Bluhmki, E., tern Med, 164, 2005–11. Wilhelm, M. and Hamilton, S. (2004) Association of 60. Malmberg, K., Ryden, L., Efendic, S., Herlitz, J., Nicol, outcome with early stroke treatment: pooled analysis of P., Waldenstrom, A., Wedel, H. and Welin, L. (1995) ATLANTIS, ECASS, and NINDS rt-PA stroke trials. Randomized trial of insulin-glucose infusion followed Lancet, 363, 768–74. by subcutaneous insulin treatment in diabetic patients 50. Wahlgren, N., Ahmed, N., Davalos, A., Ford, G. with acute myocardial infarction (DIGAMI study): ef- A., Grond, M., Hacke, W., Hennerici, M. G., Kaste, fects on mortality at 1 year. J Am Coll Cardiol, 26, M., Kuelkens, S., Larrue, V., Lees, K. R., Roine, R. 57–65. O., Soinne, L., Toni, D. and Vanhooren, G. (2007) 61. Malmberg, K., Ryden, L., Wedel, H., Birkeland, K., Thrombolysis with alteplase for acute ischaemic Bootsma, A., Dickstein, K., Efendic, S., Fisher, M., stroke in the Safe Implementation of Thrombolysis Hamsten, A., Herlitz, J., Hildebrandt, P., Macleod, K., in Stroke-Monitoring Study (SITS-MOST): an Laakso, M., Torp-Pedersen, C. and Waldenstrom, A. observational study. Lancet, 369, 275–82. (2005) Intense metabolic control by means of insulin 51. Clinical Effectiveness Evaluation Unit. (2006) National in patients with diabetes mellitus and acute myocar- Sentinel Audit of Stroke, 2006. Royal College of Physi- dial infarction (DIGAMI 2): effects on mortality and cians, London. morbidity. Eur Heart J, 26, 650–61. 52. Johnston, S. C., Fung, L. H., Gillum, L. A., Smith, 62. Gray, C. S., Hildreth, A. J., Alberti, G. K. and W. S., Brass, L. M., Lichtman, J. H. and Brown, O’Connell, J. E. (2004) Poststroke hyperglycemia: nat- A. N. (2001) Utilization of intravenous tissue-type ural history and immediate management. Stroke, 35, plasminogen activator for ischemic stroke at academic 122–6. medical centers: the influence of ethnicity. Stroke, 32, 63. Gray, C. S., Hildreth, A. J., Sandercock, P. A., 1061–8. O’Connell, J. E., Johnston, D. E., Cartlidge, N. E., 53. Els, T., Klisch, J., Orszagh, M., Hetzel, A., Bamford, J. M., James, O. F. and Alberti, K. G. Schulte-Monting, J., Schumacher, M. and Lucking, (2007) Glucose-potassium-insulin infusions in the C. H. (2002) Hyperglycemia in patients with focal management of post-stroke hyperglycaemia: the UK cerebral ischemia after intravenous thrombolysis: Glucose Insulin in Stroke Trial (GIST-UK). Lancet influence on clinical outcome and infarct size. Neurol, 6, 397–406. Cerebrovasc Dis, 13, 89–94. 64. Van Den Berghe, G., Wouters, P., Weekers, F., Ver- 54. Bruno, A., Levine, S. R., Frankel, M. R., Brott, T. waest, C., Bruyninckx, F., Schetz, M., Vlasselaers, D., G., Lin, Y., Tilley, B. C., Lyden, P. D., Broderick, Ferdinande, P., Lauwers, P. and Bouillon, R. (2001) J. P., Kwiatkowski, T. G. and Fineberg, S. E. (2002) Intensive insulin therapy in the critically ill patients. N Admission glucose level and clinical outcomes in the Engl J Med, 345, 1359–67. NINDS rt-PA Stroke Trial. Neurology, 59, 669–74. 65. Smith, D., Pernet, A., Hallett, W. A., Bingham, E., 55. Ribo, M., Molina, C., Montaner, J., Rubiera, M., Marsden, P. K. and Amiel, S. A. (2003) Lactate: a Delgado-Mederos, R., Arenillas, J. F., Quintana, M. preferred fuel for human brain metabolism in vivo. J and Alvarez-Sabin, J. (2005) Acute hyperglycemia Cereb Blood Flow Metab, 23, 658–64. 6.6 CONCLUSIONS 91

66. Bruno, A., Kent, T. A., Coull, B. M., Shankar, R. R., 74. Dahlof, B., Lindholm, L. H., Hansson, L., Schersten, Saha, C., Becker, K. J., Kissela, B. M. and Williams, B., Ekbom, T. and Wester, P. O. (1991) Morbidity L. S. (2008) Treatment of Hyperglycemia In Ischemic and mortality in the Swedish Trial in Old Patients Stroke (THIS). A randomized pilot trial. Stroke, 39, with Hypertension (STOP-Hypertension). Lancet, 338, 384–9. 1281–5. 67. Ryden, L., Standl, E., Bartnik, M., Van Den Berghe, 75. SHEP Co-Operative Research Group. (1991) Preven- G., Betteridge, J., De Boer, M. J., Cosentino, F., Jon- tion of stroke by antihypertensive drug treatment in sson, B., Laakso, M., Malmberg, K., Priori, S., Oster- older persons with isolated systolic hypertension. Fi- gren, J., Tuomilehto, J., Thrainsdottir, I., Vanhorebeek, nal results of the Systolic Hypertension in the Elderly I., Stramba-Badiale, M., Lindgren, P., Qiao, Q., Pri- Program (SHEP). JAMA, 265, 3255–64. ori, S. G., Blanc, J. J., Budaj, A., Camm, J., Dean, 76. Williams, B., Poulter, N. R., Brown, M. J., Davis, V., Deckers, J., Dickstein, K., Lekakis, J., McGregor, M.,McInnes,G.T.,Potter,J.F.,Sever,P.S.and K., Metra, M., Morais, J., Osterspey, A., Tamargo, J., McInnes, G. T. (2004) Guidelines for management of Zamorano, J. L., Deckers, J. W., Bertrand, M., Char- hypertension: report of the fourth working party of the bonnel, B., Erdmann, E., Ferrannini, E., Flyvbjerg, A., British Hypertension Society, 2004-BHS IV. JHum Gohlke, H., Juanatey, J. R., Graham, I., Monteiro, P. Hypertens, 18, 139–85. F., Parhofer, K., Pyorala, K., Raz, I., Schernthaner, 77. Adler, A. I., Stratton, I. M., Neil, H. A., Yudkin, J. S., G., Volpe, M. and Wood, D. (2007) Guidelines on Matthews,D.R.,Cull,C.A.,Wright,A.D.,Turner,R. diabetes, pre-diabetes, and cardiovascular diseases: ex- C. and Holman, R. R. (2000) Association of systolic ecutive summary. The Task Force on Diabetes and blood pressure with macrovascular and microvascular Cardiovascular Diseases of the European Society of complications of type 2 diabetes (UKPDS 36): prospec- Cardiology (ESC) and of the European Association tive observational study. Br Med J, 321, 412–19. for the Study of Diabetes (EASD). Eur Heart J, 28, 78. Dahlof, B., Sever, P. S., Poulter, N. R., Wedel, H., 88–136. Beevers, D. G., Caulfield, M., Collins, R., Kjeldsen, 68. Prospective Studies Collaboration. (1995) Cholesterol, S. E., Kristinsson, A., McInnes, G. T., Mehlsen, J., diastolic blood pressure and stroke: 13,000 strokes in Nieminen, M., O’Brien, E. and Ostergren, J. (2005) 450,000 people in 45 prospective cohorts. Lancet, 346, Prevention of cardiovascular events with an antihyper- 1647–53. tensive regimen of amlodipine adding perindopril as 69. Eastern Stroke and Coronary Heart Disease Collabo- required versus atenolol adding bendroflumethiazide as rative Research Group. (1998) Blood pressure, choles- required, in the Anglo-Scandinavian Cardiac Outcomes terol and stroke in eastern Asia. Lancet, 352, 1801–7. Trial-Blood Pressure Lowering Arm (ASCOT-BPLA): 70. Rodgers, A., MacMahon, S., Gamble, G., Slattery, J., Sandercock, P. and Warlow, C. (1996) Blood pressure a multicentre randomised controlled trial. Lancet, 366, and risk of stroke in patients with cerebrovascular dis- 895–906. ease. The United Kingdom Transient Ischaemic Attack 79. Poulter, N. R., Wedel, H., Dahlof, B., Sever, P. Collaborative Group. Br Med J, 313, 147. S., Beevers, D. G., Caulfield, M., Kjeldsen, S. E., 71. Neal, B., MacMahon, S. and Chapman, S. (2000) Blood Kristinsson, A., McInnes, G. T., Mehlsen, J., Niem- pressure lowering treatment trialists’ collaboration. Ef- inen, M., O’Brien, E., Ostergren, J. and Pocock, S. fects of ACE inhibitors, calcium antagonists and other (2005) Role of blood pressure and other variables blood pressure lowering drugs: results of prospectively in the differential cardiovascular event rates noted in designed overviews of randomised controlled trials. the Anglo-Scandinavian Cardiac Outcomes Trial-Blood Lancet, 356, 1955–64. Pressure Lowering Arm (ASCOT-BPLA). Lancet, 366, 72. Gueyffier, F., Boissel, J. P., Boutitie, F., Pocock, S., 907–13. Coope, J., Cutler, J., Ekbom, T., Fagard, R., Friedman, 80. Hope Study Investigators. (2000) Effects of an L., Kerlikowske, K., Perry, M., Prineas, R. and Schron, angiotensin-converting enzyme inhibitor, ramipril, on E. (1997) Effect of antihypertensive treatment in pa- cardiovascular events in high-risk patients. N Engl J tients having already suffered from stroke. Gathering Med, 342, 145–53. the evidence. The INDANA (INdividual Data ANalysis 81. Hope Study Investigators. (2000) Effects of ramipril on of Antihypertensive intervention trials) Project Collab- cardiovascular and microvascular outcomes in people orators. Stroke, 28, 2557–62. with diabetes mellitus: results of the HOPE study and 73. United Kingdom Prospective Diabetes Study Group. MICRO-HOPE. Lancet, 358, 1033–41. (1998) Tight blood pressure control and risk of 82. Bosch, J., Yusuf, S., Pogue, J., Sleight, P., Lonn, macrovascular and microvascular complications in E., Rangoonwala, B., Davies, R., Ostergren, J. and type 2 diabetes: UKPDS 38. Br Med J, 317, 703–13. Probstfield, J. (2002) Use of ramipril in preventing 92 CH 06 STROKE AND DIABETES

stroke: double blind randomised trial. Br Med J, 324, levels and six-year mortality from stroke in 350,977 699–702. men screened for the multiple risk factor intervention 83. Svensson, P., De Faire, U., Sleight, P., Yusuf, S. and trial. N Engl J Med, 320, 904–10. Ostergren, J. (2001) Comparative effects of ramipril 91. Yano, K., Reed, D. M. and Maclean, C. J. (1989) Serum on ambulatory and office blood pressures: a HOPE cholesterol and hemorrhagic stroke in the Honolulu Substudy. Hypertension, 38, E28–32. Heart Program. Stroke, 20, 1460–5. 84. Progress Collaborative Group. (2001) Randomised trial 92. Bucher, H. C., Griffith, L. E. and Guyatt, G. H. (1998) of a perindopril-based blood pressure-lowering regimen Effect of HMGcoA reductase inhibitors on stroke. A among 6,105 individuals with previous stroke or tran- meta-analysis of randomized, controlled trials. Ann sient ischaemic attack. Lancet, 358, 1033–41. Intern Med, 128, 89–95. 85. Patel, A., MacMahon, S., Chalmers, J., Neal, B., Wood- 93. Kearney, P. M., Blackwell, L., Collins, R., Keech, ward, M., Billot, L., Harrap, S., Poulter, N., Marre, M., A., Simes, J., Peto, R., Armitage, J. and Baigent, Cooper, M., Glasziou, P., Grobbee, D. E., Hamet, P., C. (2008) Efficacy of cholesterol-lowering therapy in Heller, S., Liu, L. S., Mancia, G., Mogensen, C. E., Pan, 18,686 people with diabetes in 14 randomised trials of C. Y., Rodgers, A. and Williams, B. (2007) Effects of statins: a meta-analysis. Lancet, 371, 117–25. a fixed combination of perindopril and indapamide on 94. Antiplatelet Trialists’ Collaboration. (1994) Collabora- macrovascular and microvascular outcomes in patients tive overview of randomised trials of antiplatelet ther- with type 2 diabetes mellitus (the ADVANCE trial): a apy - I: Prevention of death, myocardial infarction and randomised controlled trial. Lancet, 370, 829–40. stroke by prolonged antiplatelet therapy in various cat- 86. Hansson, L., Zanchetti, A., Carruthers, S. G., Dahlof, egories of patients. Br Med J, 308, 81–106. B., Elmfeldt, D., Julius, S., Menard, J., Rahn, K. 95. Halkes, P. H., Van Gijn, J., Kappelle, L. J., Koudstaal, H., Wedel, H. and Westerling, S. (1998) Effects of P. J. and Algra, A. (2006) Aspirin plus dipyridamole intensive blood-pressure lowering and low-dose aspirin versus aspirin alone after cerebral ischaemia of arterial in patients with hypertension: principal results of the origin (ESPRIT): randomised controlled trial. Lancet, Hypertension Optimal Treatment (HOT) randomised 367, 1665–73. trial. HOT Study Group. Lancet, 351, 1755–62. 96. National Institute for Health and Clinical Excellence. 87. Buse, J. B., Bigger, J. T., Byington, R. P., Cooper, (2005) Technology Appraisal 90. Clopidogrel and mod- L. S., Cushman, W. C., Friedewald, W. T., Genuth, ified release dipyridamole in the prevention of occlu- S., Gerstein, H. C., Ginsberg, H. N., Goff, D. C., Jr., sive vascular events. NICE, London. Grimm, R. H., Jr., Margolis, K. L., Probstfield, J. L., 97. Diener, H. C., Bogousslavsky, J., Brass, L. M., Simons-Morton, D. G. and Sullivan, M. D. (2007) Cimminiello, C., Csiba, L., Kaste, M., Leys, D., Action to Control Cardiovascular Risk in Diabetes Matias-Guiu, J. and Rupprecht, H. J. (2004) Aspirin (ACCORD) trial: design and methods. Am J Cardiol, and clopidogrel compared with clopidogrel alone 99, 21i–33i. after recent ischaemic stroke or transient ischaemic 88. O’Connell, J. E. and Gray, C. S. (1996) Treatment attack in high-risk patients (MATCH): randomised, of post-stroke hypertension. A practical guide. Drugs double-blind, placebo-controlled trial. Lancet, 364, Aging, 8, 408–15. 331–7. 89. Robinson, T. G. and Potter, J. F. (2004) Blood pressure 98. Gaede, P., Vedel, P., Larsen, N., Jensen, G. V., Parving, in acute stroke. Age Ageing, 33, 6–12. H. H. and Pedersen, O. (2003) Multifactorial interven- 90. Iso, H., Jacobs, D. R., Jr., Wentworth, D., Neaton, J. tion and cardiovascular disease in patients with type 2 D. and Cohen, J. D. (1989) Serum cholesterol diabetes. N Engl J Med, 348, 383–93. 7 Diabetes-Related Renal Disease in Older People

Latana A. Munang1 and John M. Starr2 1 Liberton Hospital, 113 Lasswade Road, Edinburgh, UK 2 University of Edinburgh, Royal Victoria Hospital, Edinburgh, UK

replacement therapy [4]. In the USA, the incidence of Key messages diabetic end-stage renal failure is 153 cases per million population [5]. • Diabetic renal disease is associated with increased cardiovascular mortality and progression to end-stage renal failure. 7.2 Changes in the diabetic kidney • Early identification of renal disease through regular screening, good management of cardio- Injury in the diabetic kidney was initially thought to be vascular risk factors, tight blood pressure and due to haemodynamic changes of hyperperfusion and glycaemic control are key treatment goals. hyperfiltration, but it is increasingly clear that these • Appropriate patients should be referred promptly changes are only one part of complex pathophysiolog- to the nephrologist. ical interactions. The results of recent studies have suggested that genetic influences and metabolic pathways involving oxidative stress, endothelial dysfunction, cy- 7.1 Introduction tokines and growth factors are equally important [6–8]. Although previously debated as separate entities, it is Diabetes mellitus (DM) is the leading cause of chronic now acknowledged that the basic pathophysiological kidney disease in Western countries, with approxi- mechanisms underlying diabetic nephropathy are sim- mately 40% of patients with type 1 and type 2 DM ilar in both type 1 and type 2 DM. However, many eventually developing diabetes-related renal disease people with type 2 DM have the additional factors [1]. In addition, up to 20% of patients newly diagnosed of hypertension, obesity, dyslipidaemia and ischaemic with type 2 DM will already have diabetic renal dis- renal disease of arteriosclerosis, all of which can con- ease, and a further 30–40% will develop the condition, tribute to kidney damage, producing complex patterns mostly within 10 years [2]. of nephropathy [7]. Diabetic nephropathy is independently associated Structurally, microscopic changes in the diabetic with increased cardiovascular mortality [3] as well kidney include a thickening of the glomerular basement as with a higher risk of progression to end-stage membrane and an increase in the mesangial matrix renal failure. In the United Kingdom, diabetic renal and mesangial cell proliferation, giving rise to dif- disease accounts for 20% of people commencing renal fuse glomerulosclerosis. Nodular glomerulosclerosis is

Diabetes in Old Age Third Edition Edited by Alan J. Sinclair © 2009 John Wiley & Sons Ltd. ISBN: 978-0-470-06562-4 94 CH 07 DIABETES-RELATED RENAL DISEASE IN OLDER PEOPLE characterized by Kimmelstiel–Wilson lesions, which Microalbuminuria or proteinuria does not automati- are distinctive ball-like areas of mesangial expansion cally confirm diabetic nephropathy. Hence, it is impor- with halos of glomerular capillary loops around the tant to exclude other causes of renal disease by taking nodule. There is also podocyte effacement and loss. In a full clinical history and examination. However, in pa- the tubulointerstitium, the tubular basement membrane tients with at least a 10-year history of type 1 DM with is also thickened, accompanied by tubular atrophy, in- retinopathy, increasing albuminuria and blood pres- terstitial fibrosis and arteriosclerosis. sure, as well as a declining renal function in keeping The earliest indicator of renal disease in diabetes with the natural history of diabetic nephropathy, no is microalbuminuria. This refers to albumin values in further investigations are needed [6]. The association the urine that, although low, are above normal lev- between renal disease, retinopathy, albuminuria and els. Next, proteins appear non-selectively in the urine, duration of diabetes is less clear-cut in type 2 DM, and this is followed by decline in glomerular filtra- but in the absence of any other suspicious features tion function and ultimately, established renal failure. the deteriorating renal function can be assumed to be Approximately one-third of people with microalbumin- diabetic renal disease, without the need for further in- uria will progress to proteinuria, one-third will remain vestigations. Serological autoantibodies, renal imaging microalbuminuric, and one-third will revert to normal and kidney biopsy may be indicated when the diagnosis albumin excretion [9]. Of those who become protein- is in doubt. However, the risk:benefit ratio of undertak- uric, almost all will develop end-stage renal disease ing these investigations must be carefully considered or die prematurely of cardiovascular disease [6]. It is as these tests carry a substantial risk, particularly for therefore crucial that diabetic renal disease is detected older people. Complications of renal biopsy include as early as possible so as to allow prompt intervention. bleeding and perinephric haematomas, while radiological studies using contrast agents may precipitate an 7.3 Screening for diabetic renal acute decline in renal function. disease Screening for diabetic renal disease also includes the regular measurement of serum creatinine levels, at least annually. Serum creatinine is, however, a Current guidelines recommend annual screening for microalbuminuria and proteinuria [10–12], preferably crude measure of renal function, as levels may vary using an early morning urine sample to avoid postural depending on age, gender, muscle mass and diet. In proteinuria and calculating the albumin: creatinine older people with a relatively lower muscle mass, ratio (ACR), using a laboratory method as albumin serum creatinine levels are often within the normal concentration alone can be unreliable. Daily (24 h) range, even when their renal function has declined urine collections are cumbersome and unnecessary, as significantly, leading to under-detection [14]. spot ACR measurements have demonstrated an excel- Because of the insensitivity of serum creatinine, cur- lent correlation with these [13]. If protein is detected, rent guidelines now recommend using the four-variable any underlying urinary tract infection, intercurrent Modification of Diet in Renal Disease (MDRD) for- acute illness, congestive heart failure, severe hyperten- mula to estimate the glomerular filtration rate (GFR) sion and haematuria should be excluded as potential (Table 7.1) [15]. Although less accurate in early causes. Three positive tests are required over a period chronic kidney disease, this is the preferred method of less than 6 months in order to make a firm diagnosis in older people [16, 17]. It is also important to of persistent microalbuminuria or proteinuria. Current note that the development of the MDRD formula did UK guidelines define microalbuminuria as an ACR not include patients with diabetes [18], and conse- of >2.5 mg mmol−1 in men and >3.5 mg mmol−1 in quently potential accuracy problems persist when us- women, while proteinuria is defined as an ACR of ing this formula in this patient population. Cystatin ≥ 30 mg mmol−1 [10, 11]. ACR values in women C, a low-molecular-weight cysteine protease inhibitor are higher for the equivalent level of urinary albumin which is freely filtered by the kidney, can also be mea- excretion in men, mainly because women normally sured in the serum as a marker of kidney function, but have lower urinary creatinine concentrations. In the this has yet to achieve routine clinical use. Cystatin C USA, guidelines define microalbuminuria as an ACR has been shown to be more accurate than creatinine between 30 and 300 mg g−1, while proteinuria is an in detecting early kidney disease in patients with dia- ACR of >300 mg g−1 [12]. betes [19] and in older people, in addition to serving 7.5 MANAGEMENT OF DIABETIC RENAL DISEASE 95

Table 7.1 Estimation of the glomerular ﬁltration rate using the four-variable MDRD equation. GFR (ml min−1 per 1.73 m2) =186 ×{[serum creatinine (μmol l−1)/88.4]−1.154} × age (years)−0.023 × 0.742 if female, and × 1.21 if African-American as a strong prognostic indicator of death and cardio- GFR decline can be estimated by plotting the reciprocal vascular disease (CVD) [20]. of creatinine concentration against time; a change in the gradient of the curve, indicating an acceleration of the rate of decline, should trigger investigations 7.4 Chronic kidney disease to determine the causes of worsening renal function, which may potentially be reversible. UK guidelines for Chronic kidney disease (CKD) is deﬁned as either kid- referral to a nephrologist are outlined in Table 7.3 [15]. ney damage or a decreased GFR of <60 ml min−1 per 1.73 m2 persisting for at least 3 months, regardless of the underlying aetiology. Kidney damage may be indi- 7.5 Management of diabetic renal cated by persistent microalbuminuria or proteinuria or disease haematuria, by radiologically demonstrable structural abnormalities, or by biopsy-proven chronic glomeru- 7.5.1 Primary prevention lonephritis. The level of kidney function determines the stage of CKD (Table 7.2) [15, 21]. The ultimate aim is to intervene early enough to Early CKD causes few, if any, symptoms – which is prevent the development of diabetic renal disease. The why an initial recognition through regular screening of results of several studies have shown that, the better patients with diabetes is essential. Patients may present the glucose control, the lower the risk of developing with a poor appetite, nausea and vomiting, tiredness, microalbuminuria [22, 23]. Thus, the lowest possible breathlessness, peripheral oedema, itch, cramps or rest- HbA1c should be the target, generally below 6.5–7.5%, less legs, but they may often be asymptomatic, even at adjusted to suit each individual. Good blood pressure CKD Stage 5. control also reduces the risk of developing diabetic Kidney function tests should be carried out at least renal disease [2], with the upper limit of acceptable annually to monitor progression of the disease, and blood pressure generally 140/80 mmHg. more frequently in the latter stages of CKD. Ideally, It is important to remember that the intensive man- patients with CKD Stage 3 should be assessed every agement of older patients may be hazardous because 6 months, and those with CKD Stage 4 or 5 every hypotension and hypoglycaemia occur more frequently 3 months. The GFR of patients with diabetes and than in younger people. Both conditions can be con- proteinuria can decline as rapidly as 1.2 ml min−1 per trolled using the usual medications, but lower starting month if left untreated, although in some patients the doses should be used with careful titration and moni- GFR may remain stable for a long time [1]. The rate of toring for side effects.

Table 7.2 Classiﬁcation of chronic kidney disease [15, 21]. Stage GFR (ml min−1 per Description 1.73 m2) 1 >90 Normal kidney function but with evidence of kidney damage 2 60–89 Mildly reduced kidney function with evidence of kidney damage 3 30–59 Moderately reduced kidney function 4 15–29 Severely reduced kidney function 5 <15 or on dialysis Established renal failure 96 CH 07 DIABETES-RELATED RENAL DISEASE IN OLDER PEOPLE

Table 7.3 Criteria for referral to specialist nephrology services [15]. Stage of CKD Referral to nephrologists 1 and 2 Not required unless other problems occur (see below) 3 Routine referral if: Progressive fall in GFR Microscopic haematuria Urinary protein/creatinine ratio >45 mg mmol−1 Unexplained anaemia, abnormal potassium, calcium or phosphate Suspected systemic illness, e.g., systemic lupus erythematosus Uncontrolled blood pressure (>150/90 mmHg while receiving three antihypertensive agents) 4 Urgent referral or routine referral if stable (<2mlmin−1 per 1.73 m2 change in GFR over ≥ 6 months) 5 Immediate referral Renal problems Immediate referral irrespective of GFR Malignant hypertension Hyperkalaemia (K+>7 mmol l−1) Urgent referral Nephrotic syndrome Routine referral Dipstick proteinuria and urine protein/creatinine ratio >100 mg mmol−1 Dipstick proteinuria and microscopic haematuria Macroscopic haematuria, but urological tests negative

7.5.2 Management of microalbuminuria higher insulin doses being required [24]. However, im- and proteinuria paired kidney function also means that the half-life of insulin is prolonged because of decreased degradation, Although it may not be possible to completely pre- and this will result in more hypoglycaemic episodes. vent the development of diabetic renal disease, the Consequently, insulin doses must be titrated cautiously. process can be delayed signiﬁcantly. The rate of change Similarly, the clearance of most oral agents is also of albuminuria has been shown to be an independent decreased, giving rise to a higher risk of adverse side predictor of death and cardiovascular events in these effects. In patients with CKD Stage 3–5, long-acting patients [3]. Therefore, the main goals of treatment sulphonylureas such as chlorpropamide and gliben- concentrate on preventing progression from microal- clamide should be avoided because of the risk of pro- buminuria to proteinuria, slowing down the decline of found hypoglycaemia. Glipizide can be used without renal function in patients with proteinuria, and prevent- dose adjustment as its metabolites are inactive, while ing cardiovascular events. α-glucosidase inhibitors should also be avoided because of potential hepatic damage. Although there is a lower risk of hypoglycaemia with metformin, it is 7.5.3 Glycaemic control contraindicated in patients with serum creatinine con- − − Good glycaemic control is important. On the other centrations ≥ 1.5mgdl 1 in men and ≥ 1.4mgdl 1 in hand, the lower the HbA1c goal and the tighter the women, because potential accumulation even in mild glycaemic control, the higher the risk of more frequent renal disease places patients at risk of lactic acidosis and more severe hypoglycaemic episodes, particularly [12]. Recent studies have suggested that thiazolidine- in older patients. diones such as roziglitazone and pioglitazone may have While there are currently no evidence-based guide- a protective effect in slowing down the progression of lines in recommending which types of insulin to use, diabetic renal disease [25], and as these drugs are me- it is important to note that the effect of insulin may tabolized by the liver they are safe to use in patients be impaired in patients with renal disease, resulting in with CKD, without any dose adjustment. 7.7 MANAGEMENT OF CKD AND ITS COMPLICATIONS 97 In advanced renal disease, although good glycaemic increased susceptibility to infection and high mortality. control may no longer be necessary to prevent further It is therefore important that each patient is interviewed deterioration in kidney function, it is still important for by a dietician for individual dietary assessment and preventing the progression of retinopathy, neuropathy advice. and macrovascular disease.

7.5.4 Blood pressure and renin–angiotensin 7.6 Managing cardiovascular risk system blockade Cardiovascular disease is the leading cause of death The importance of reducing systemic blood pres- in CKD, with most patients more likely to die from sure in delaying the progression of diabetic renal a cardiovascular event before reaching end-stage re- disease is well recognized. Additionally, blocking nal failure. Aggressive risk management is therefore the renin–angiotensin system (RAS) with angiotensin- important and will also delay the progression of renal converting enzyme (ACE) inhibitors or angiotensin re- disease. Patients should be advised on smoking cessa- ceptor blockers (ARBs) is renoprotective, independent tion, weight loss, exercise, and limiting both alcohol of its blood pressure-reducing effect [26]. Therefore, and sodium intake. Aspirin is also recommended for all diabetic patients with microalbuminuria and pro- primary and secondary prevention. teinuria should be commenced on an ACE inhibitor or ARB. It is important to re-check serum creatinine and potassium at 1–2 weeks after commencing these 7.6.1 Dyslipidaemia agents, and also after each increase in dose. While a The management of dyslipidaemia is important in small rise in serum creatinine is common, a creatinine reducing the cardiovascular risk of all patients with increase of >20% or a fall in GFR of >15% should diabetic renal disease. It is suggested that lipid prompt consideration of withdrawal and further inves- levels be measured annually, aiming for a target tigation for renal artery stenosis. LDL-cholesterol of <100 mg dl−1 for CKD Stages The systemic blood pressure will often remain 1–4 [12]. In addition to lifestyle measures such as elevated, despite the maximum tolerated dose of ACE diet, exercise and weight loss, the use of statins inhibitor or ARB, and it is not uncommon for patients are recommended as the agents of choice, although with advancing renal disease to require four or five combination therapies may be required to achieve antihypertensive agents, including calcium-channel target levels [27]. blockers, beta-blockers, alpha-blockers and diuretics. Loop diuretics are preferred to thiazides in the later stages of CKD, as the diuretic efficacy of thiazides 7.7 Management of CKD and its tends to decrease with deteriorating renal function complications and the potential for electrolyte disturbances becomes higher. The target blood pressure varies between Most patients up to CKD Stage 3 can be man- guidelines, but range from 120/70 to 140/80 mmHg aged safely by general practitioners in the community. [6, 10–12]. The main complications of CKD involve the bones, anaemia and metabolic acidosis [28]. In addition to the above-described measures, all patients should un- 7.5.5 Diet and nutrition dergo at least annual monitoring of their haemoglobin The reduction of dietary protein intake to 0.8 g kg−1 (Hb), potassium, calcium and phosphate. Renal ultra- body weight is recommended in CKD Stages 1–4, in sonography may be indicated in patients with lower order to reduce proteinuria and slow the rate of renal urinary tract symptoms, persistent hypertension and un- function decline [12]. However, patients with CKD are expected progressive declining renal function. already subject to dietary restrictions because of hy- Renal bone disease starts at Stage 3, causing an perkalaemia, hyperphosphataemia and hypertension, in impaired renal production of active 1,25-dihydroxy addition to the restrictions already required because of vitamin D. This is very important in older people, as diabetes. Malnutrition may occur as a consequence of a large proportion will have coexisting osteoporosis. these imposed restrictions, and this can lead to pro- Both, hypocalcaemia and hyperphosphataemia cause gressive muscle weakness, poor exercise tolerance, an secondary hyperparathyroidism, and treatment with 98 CH 07 DIABETES-RELATED RENAL DISEASE IN OLDER PEOPLE ergocalciferol or cholecalciferol with calcium supple- patient’s blood into a dialyzer, where molecules dif- ment may be indicated. Phosphate binders can also be fuse across a semi-permeable membrane and fluid used for treating hyperphosphataemia. is removed by ultrafiltration. Anticoagulation with Anaemia of CKD begins at Stage 3 and, if left heparin and good vascular access are required for untreated, may cause left ventricular hypertrophy and haemodialysis. Ideally, a subcutaneous arteriovenous failure, a poor quality of life and impaired cognition. fistula should be formed 6 months before starting dial- Anaemic patients with Hb <11gdl−1 should be con- ysis, but synthetic grafts, central venous catheters and sidered for treatment with erythropoiesis-stimulating semi-permanent tunnelled central catheters may be also agents and intravenous iron [29]. A stable Hb between be used in the short to intermediate term. Generally, pa- 10.5 and 12.5 g dl−1 is recommended, depending on tients undergo three haemodialysis sessions per week, the individual’s functional needs. Blood transfusions each lasting 3–5 h. The procedure is normally carried should be avoided if possible in those patients where out in hospital units, although home haemodialysis may renal transplantation is a treatment option. be possible for appropriate patients, with good out- Metabolic acidosis is uncommon before CKD Stage comes [34]. 4, but can contribute to bone disease, muscle wasting, anorexia, hypoalbuminaemia and a progressive deteri- 7.7.3 Peritoneal dialysis oration of renal function. It should be corrected with Peritoneal dialysis is normally carried out at home, thus supplementary sodium bicarbonate, aiming to maintain allowing much greater patient independence. However, ≥ −1 plasma bicarbonate at 20 mmol l . because it uses the peritoneum as the semi-permeable All patients who reach CKD Stage 4 or 5 should membrane for dialyzing, it is contraindicated in pa- be discussed with a nephrologist, because they must tients who have had major abdominal surgery or peri- be adequately informed about their treatment options toneal adhesions, unrepaired inguinal hernias or comas they approach end-stage renal failure. Most patients promised respiratory function; thus, a large number of need at least one year to prepare for renal replacement older people are excluded. Common problems include therapy (RRT) [30], and a prompt referral is associated infection (hence, good personal hygiene is essential) with better outcomes [31]. Most patients who reach and fluid overload due to inadequate ultrafiltration. end-stage renal failure have progressed from earlier stages of CKD, yet up to one-third are referred late and 7.7.4 Renal transplantation will only be examined by a nephrologist less than one month before requiring RRT. Delays in these referrals Renal transplantation is clearly preferable to dialysis as occur more frequently in older people, contributing to it offers CKD patients a return to normality. However, excess mortality [32]. the scarcity of donor organs means that many older patients are not considered for transplantation because of their comorbidity and shorter life expectancy. There 7.7.1 End-stage renal disease and RRT is currently no age limit to transplantation in the UK, Today, increasing numbers of older people are under- but only 14% of renal transplant recipients are over going RRT. In the UK, the median age of patients start- the age of 60 [35], despite increasing evidence showing dialysis has risen from 63.9 years in 1998 to 65.5 ing transplantation to be worthwhile in older patients years in 2005, and at present 25% of dialysis patients if they are selected carefully and their immunosuppres- are aged >75 years [4]. The guidelines recommend sion tailored sensibly [36, 37]. The main complications starting RRT when the GFR falls below 15 ml min−1 are graft rejection, infection and malignancy, but CVD per 1.73 m2 and there are problems of symptomatic remains the leading cause of death. uraemia, uncontrollable fluid balance, blood pressure or malnutrition. Even when asymptomatic, RRT should be started when the GFR falls to <6mlmin−1 per 7.8 Conservative management of 1.73 m2 [33]. renal disease

Although life-prolonging, RRT may not be the best 7.7.2 Haemodialysis treatment for all patients. Survival after dialysis for Haemodialysis is the most common form of RRT older, heavily dependent patients with more comor- employed in older people. This involves passing the bidity is not signiﬁcantly longer than those treated 7.9 CONCLUSIONS 99 palliatively [38]. Neither is the patient’s quality of Today, conservative care specialist nurses are available life necessarily better, as many older patients receiv- at many renal units to facilitate this treatment option ing RRT spend a considerable proportion of their time for patients. in hospital with a higher risk of developing multiple complications [39]. Conservative or non-dialytic man- 7.9 Conclusions agement aims to relieve the symptoms of end-stage renal failure, including the treatment of anaemia with The current aging population and global epidemic erythropoiesis-stimulating agents and the control of of diabetes means that the number of people with nausea and pruritus with appropriate drugs. End of diabetic renal disease will continue to increase. Early life care is equally important, and requires a close col- recognition, aggressive management of cardiovascular laboration with palliative services and primary care. risk factors and tight control of both blood pressure and

Prevention

Good glycaemic control, target HbA1c 6.5–7.5% Good blood pressure control, target 140/80 mmHg

Screening Annual creatinine Exclude acute renal failure or estimated GFR Low eGFR Exclude outflow obstruction (eGFR) Review medications Re-check creatinine

Annual urinalysis Protein present Repeat positive tests twice in 6 months Exclude infection and other causes Calculate albumin:creatinine ratio (ACR) - Microalbuminuria: ACR >2.5mgmmol−1 in men >3.5mgmmol−1 in women Proteinuria: ACR≥30mgmmol−1

Established renal disease Glycaemic control Titrate insulin doses carefully Avoid long-acting sulphonylureas Avoid metformin in advanced CKD

Blood pressure control Start ACE inhibitor or ARB Re-check creatinine and potassium

Dyslipidaemia Lifestyle measures, consider statin Aim for LDL-cholesterol <100mg dl−1

Cardiovascular risk Commence aspirin Stop smoking Weight loss and exercise Limit alcohol and sodium intake

Nutrition Dietary assessment and advice

Complications of CKD Check haemoglobin, potassium, calcium, phosphate, parathyroid hormone regularly Bone Consider ergocalciferol/cholecalciferol with calcium supplements and phosphate binders Anaemia Consider erythropoeisis-stimulating agents/intravenous iron Acidosis Consider sodium bicarbonate

Disease progression Monitor eGFR regularly Haemodialysis Refer CKD Stage 4–5 to nephrologist Peritoneal dialysis Prepare for end-stage renal failure Renal transplantation Conservative care

Figure 7.1 Approach to management of diabetic renal disease. 100 CH 07 DIABETES-RELATED RENAL DISEASE IN OLDER PEOPLE

glycaemia are key strategies in reducing mortality and Diabetes and Chronic Kidney Disease. Am J Kidney delaying progression to end-stage renal failure (Figure Dis 2007; 49 (Suppl. 2): S1–S180. 7.1). Patients should be informed at an early stage 13. Gansevoort RT, Verhave JC, Hillege HL, Burgerhof JG, regarding the likely course of their disease, and their Bakker SJ, de Zeeuw D and de Jong PE. The validity prognosis. Appropriate patients should be referred to of screening based on spot morning urine samples to nephrology services promptly, so that they may be detect subjects with microalbuminuria in the general adequately prepared to make informed choices about population. Kidney Int 2005; 98 (Suppl.): S28–35. their treatment options. 14. Swedko PJ, Clark HD, Paramsothy K and Akbari A. Serum creatinine is an inadequate screening test for renal failure in elderly patients. Arch Intern Med 2003; 163: 356–60. References 15. Royal College of Physicians. Chronic kidney disease in adults: UK guidelines for identification, management 1. Gross JL, de Azevedo MJ, Silveiro SP, Canani LH, and referral. Royal College of Physicians, London; Caramori ML and Zelmanovitz T. Diabetic nephropa- 2005. thy: diagnosis, prevention, and treatment. Diabetes 16. Burkhardt H, Hahn T, Gretz N and Gladisch R. Bedside Care 2005; 28 (1): 164–76. estimation of the glomerular filtration rate in hospital- 2. Thomas MC and Atkins RC. Blood pressure lowering ized elderly patients. Nephron 2005; 101: c1–8. for the prevention and treatment of diabetic kidney 17. Cirillo M, Anastasio P and De Santo NG. Relationship disease. Drugs 2006; 66 (17): 2213–34. of gender, age and body mass index to errors in 3. Yuyun MF, Dinneen SF, Edwards OM, Wood E and predicted kidney function. Nephrol Dial Transplant Wareham NJ. Absolute level and rate of change of 2005; 20: 1791–8. albuminuria over 1 year independently predict mortal- 18. Poggio ED, Wang X, Greene T, Van Lente F and Hall ity and cardiovascular events in patients with diabetic PM. Performance of the modification of diet in renal nephropathy. Diabet Med 2003; 20: 277–82. disease and Cockcroft-Gault equations in the estimation 4. Ansell D, Feest TG, Tomson C, Williams AJ and War- wick G. UK Renal Registry Ninth Annual Report, of GFR in health and in chronic kidney disease JAm December. The Renal Association, Petersfield, Hamp- Soc Nephrol 2005; 16: 459–66. shire; 2006. 19. MacIsaac RJ, Tsalamandris C, Thomas MC, Pre- 5. U.S. Renal Data System. Annual Data Report: Atlas of maratne E, Panagiotopoulos S, Smith TJ, Poon A, Chronic Kidney Disease and End-Stage Renal Disease Jenkins MA, Ratnaike SI, Power DA and Jerums in the United States. National Institutes of Health, Na- G. The accuracy of cystatin C and commonly used tional Institute of Diabetes and Digestive and Kidney creatinine-based methods for detecting moderate and Diseases, Bethesda; 2007. mild chronic kidney disease in diabetes Diabet Med 6. Marshall SM. Recent advances in diabetic nephropathy. 2007; 24: 443–8. Postgrad Med J 2004; 80: 624–33. 20. Shlipak MG, Wassel Fyr CL, Chertow GM, Harris TB, 7. Wolf G. New insights into the pathophysiology of dia- Kritchevsky SB, Tylavsky FA, Satterfield S, Cummings betic nephropathy: from haemodynamics to molecular SR, Newman AB and Fried LF. Cystatin C and mor- pathology. European Journal of Clinical Investigation tality risk in the elderly: the health, aging, and body 2004; 34: 785–96. composition study. J Am Soc Nephrol 2006; 17 (1): 8. Caramori ML and Mauer M. Diabetes and nephropathy. 254–61. Curr Opin Nephrol Hypertens 2003; 12: 273–82. 21. Levey AS, Coresh J, Balk E, Kausz AT, Levin A, 9. Marshall SM and Flyvbjerg A. Prevention and early Steffes MW, Hogg RJ, Perrone RD, Lau J and Eknoyan detection of vascular complications of diabetes. Br Med G. National Kidney Foundation practice guidelines for J 2006; 333: 475–80. chronic kidney disease: Evaluation, classification and 10. National Institute for Clinical Excellence. Management stratification. Ann Intern Med 2003; 139 (2): 137–47. of Type 2 diabetes: the prevention and early manage- 22. Diabetes Control and Complications Trial (DCCT) Re- ment of renal disease. National Institute for Clinical search Group. The effect of intensive treatment of dia- Excellence, London; 2002. betes on the development and progression of long term 11. Scottish Intercollegiate Guidelines Network. Manage- complications in insulin-dependent diabetics. N Engl J ment of diabetes. SIGN Publication No. 55, November Med 1993; 329: 977–86. 2001. 23. UK Prospective Diabetes Study. Intensive blood glu- 12. National Kidney Foundation. KDOQI Clinical Practice cose control with sulphonylureas or insulin compared Guidelines and Clinical Practice Recommendations for with conventional treatment and risk of complications 7.9 CONCLUSIONS 101

in patients with type 2 diabetes. Lancet 1998; 352: of poor outcome in the very elderly dialysis patient. 837–53. Nephrol Dial Transplant 2006; 21: 962–7. 24. Cavanaugh KL. Diabetes management issues for pa- 33. European Renal Association. European Best Practice tients with chronic kidney disease. Clinical Diabetes Guidelines. Section I: Measurement of renal function, 2007; 25 (3), 90–7. when to refer and when to start dialysis: I.3 When 25. Saraﬁdis PA and Bakris GL. Protection of the kidneys to start dialysis. Nephrol Dial Transplant 2002; 17, by thiazolidinediones: an assessment from bench to 10–11. bedside. Kidney Int 2006; 70: 1223–33. 34. Saner E, Nitsch D, Descoeudres C, Frey FJ and 26. Strippoli GFM, Bonifati C, Craig M, Navaneethan Uehlinger DE. Outcome of home haemodialysis pa- SD and Craig JC. Angiotensin converting enzyme in- tients: a case cohort study. Nephrol Dial Transplant hibitors and angiotensin II receptor antagonists for 2005; 20 (3): 604–10. preventing the progression of diabetic kidney disease. 35. NHS Blood and Transplant. UK Transplant Activity Cochrane Database of Systematic Reviews 2006; (4), Report 2006-2007. NHS, Bristol, UK; 2007 CD006257. 36. Segoloni GP, Messina M, Giraudi R, Leonardi G, 27. Solano MP and Goldberg RB. Management of dyslipi- Torta E, Gabrielli D, Ferrari A, Pellu V, Tattoli F demia in diabetes. Cardiology in Review 2006; 14 (3); and Fop F. Renal transplantation in patients over 65 125–35. years of age: no more a contraindication but a growing 28. Lederer E and Ouseph R. Chronic kidney disease. Am indication. Transplantation Proceedings 2005; 37 (2): J Kidney Dis 2007; 49 (1): 162–71. 721–5. 29. National Collaborating Centre for Chronic Conditions. 37. Otero-Ravina˜ F, Rodr´ıguez-Mart´ınez M, Gude F, Anaemia management in chronic kidney disease: na- Gonzalez-Juanatey JR, Valdes F and Sanchez-Guisande tional clinical guideline for management in adults and D. Renal transplantation in the elderly: does patient children. Royal College of Physicians, London, 2006. age determine the results? Age and Ageing 2005; 34: 30. Department of Health. National Service Framework for 583–7. Renal Services: Part 2 - Chronic kidney disease, acute 38. Smith C, Da Silva-Gane M, Chadna S, Warwicker P, renal failure and end of life care. Department of Health, Greenwood R and Farrington K. Choosing not to dial- London; 2005. yse: evaluation of planned non-dialytic management 31. Stack AG. Impact of timing of nephrology referral and in a cohort of patients with end-stage renal failure. pre-ESRD care on mortality risk among new patients Nephron Clin Pract 2003; 95: c40–6. in the United States. Am J Kidney Dis 2003; 41: 39. Munshi SK, Vijayakumar N, Taub NA, Bhullar H, Lo 310–18. TCN and Warwick G. Outcome of renal replacement 32. Schwenger V, Morath C, Hofmann A, Hoffmann O, therapy in the very elderly. Nephrol Dial Transplant Zeier M and Ritz E. Late referral – a major cause 2001; 16: 128–33. 8 Management of Eye Disease and Visual Loss

Nina Tumosa Geriatrics Research, Education and Clinical Center, St Louis VAMC, and Division of Geriatrics, Saint Louis University, St Louis MO, USA

complications are attributed to chronic hyperglycemia, Key messages vascular damage and leakage, oedema, capillary basement membrane thickening, neovascularization, haem- • The management of diabetic retinopathy re- orrhage and ischaemia. Diabetic retinopathy is believed quires good communication between the patient to be a leading cause of blindness in the industrialized and his/her health care providers. world in people between the ages of 25 and 74 [1], • Diabetic retinopathy can result in blindness or and the fourth leading cause of blindness in people of severe visual impairment if diabetes mellitus all ages in developing countries [2]. The prevalence (DM) is left untreated. of DM for all age-groups worldwide was estimated to • While new treatments such as the use of antibodies, cytokines and fusion proteins to prevent be 2.8% in 2000 and 4.4% in 2030. The total num- diabetic retinopathy show promise, at present ber of people with DM is projected to rise from 171 the best way to prevent diabetic retinopathy is million in 2000 to 366 million in 2030. The most im- strict control of DM through the use of nutritional portant demographic change to DM prevalence across therapy, exercise and pharmacological agents. the world appears to be the increase in the proportion of people aged over 65 years [3]. Diabetic retinopathy causes considerable morbidity and mortality, and has a 8.1 Introduction large economic impact [4]; consequently, its prevention and successful treatment are desirable public health Diabetes mellitus (DM) is a chronic disease with a goals [5]. long-term risk for visual and neurological impairments. Diabetic retinopathy (retinal changes that occur in patients with DM) is characterized by microaneurysms, 8.2 Risk factors exudates, haemorrhages and, less commonly, neovascularization, and is more common than other microvas- Many risk factors may affect the rate at which diabetic cular complications of DM. Diabetic patients with retinopathy progresses (see Table 8.1). It is incumbent poorer long-term glycemic control are more vulner- upon the health care team to educate the diabetic able to diabetic retinopathy than are those with better patient about these risk factors and to develop a care glycemic control. Retinopathy and other microvascular plan that addresses those factors that are modiﬁable

Table 8.1 Risk factors for diabetic retinopathy. The ocular assessment that optometrists or ophthalmologists perform may include tests for visual acu- Risk factor Reference(s) ity, pupil function assessment, extraocular muscle Age [6, 7] motilities, visual field distortions (Amsler Grid), colour Race [8–10] vision changes, tear film break-up time and contrast Obesity [8] Smoking [11] sensitivity changes, as well as the performance of Proteinuria [12] biomicroscopy, tonometry, gonioscopy, dilated stereo- Depression [13, 14] scopic fundus examination and fundus photography Dyslipidaemia [11] [22]. The clinical signs that are indicative of the pres- Duration of DM [6, 7, 9, 10] ence of diabetic retinopathy, and which can be con- Eating disorders [15] firmed during this thorough visual assessment of the Gestational diabetes [16] Poor glycemic control [6, 17] retina include any of the following items: Parental history of DM [8] • Cotton wool spots Cardiovascular disease [18, 19] Uncontrolled systemic hypertension [9, 11, 12] • Arteriolar narrowing Rapid pre-cataract surgery glycemic control [20] • Fibrotic proliferation • Decreased visual acuity in an effort to prevent or slow down the onset or • Hard exudates (Figure 8.1) progression of diabetic retinopathy. There are many clinical signs that are indicative of • Colour vision abnormalities the presence of diabetic retinopathy. Optometrists and • Blot haemorrhages (Figures 8.2 and 8.3) ophthalmologists are the most likely providers to first • find or confirm the diagnosis of diabetic retinopathy Flame-shaped haemorrhages (Figure 8.2) to their diabetic patients. These providers perform • Venous tortuosity and beading thorough ocular assessments on patients who report • Metamorphosia and/or scotomas a history of having been diagnosed with DM or who are suspected of having DM because of the presence of • Dot haemorrhages and microaneurysms multiple risk factors listed above, or because of specific • Intraretinal microvascular abnormalities visual complaints such as fluctuating vision or ‘seeing red’ (which is a sign of a vitreous haemorrhage). Macular changes that are secondary to the presence of The American Diabetes Association’s Clinical Practice diabetic retinopathy and can best be confirmed during Guidelines [21] for examining a patient with diabetic a dilated fundus examination include: retinopathy include performing: Retinal detachment Vitreous haemorrhage • Comprehensive eye examinations within 3–5 years Preretinal haemorrhage of the onset of DM (type 1 diabetics) and immedi- Disc neovascularization (Figure 8.4) ately after initial diagnosis (type 2 diabetics). Neovascularization elsewhere (Figure 8.5). • Subsequent annual dilated fundus examinations. • Comprehensive eye examinations for pregnant 8.3 Management of diabetic women in the first trimester with close follow-up retinopathy and counselling throughout pregnancy and in the first year post-partum. The management of diabetic retinopathy requires the • Prompt referral to a retinal specialist of patients with attention of all members of the health care team. macular oedema, severe non-proliferative diabetic The team should include the primary care physician, nurse, social worker, nutritionist, exercise therapist, op- retinopathy or proliferative diabetic retinopathy. tometrist, ophthalmologist, endocrinologist, health ed- • Strict control of elevated blood pressure. ucator and geriatrician, as well as the family caregiver 8.3 MANAGEMENT OF DIABETIC RETINOPATHY 105

Figure 8.1 Maculopathy. Certain pathological changes occur in the macula as a result of early diabetes mellitus. These changes include leakage of blood vessels around the macula (black space in the middle of the picture). Remnants of blood products show as white spots (hard exudates) arrayed around the macula.

Figure 8.2 Background retinopathy. Pathological changes that occur early in diabetes mellitus elsewhere in the retina (other than the macula, as shown in Figure 8.1) include haemorrhages at the ends of the blood vessels. These appear either as dark circles to the left of the centrally located macula (blot haemorrhages) or as a more diffuse pooling of blood (ﬂame haemorrhages) in the superior and inferior quadrants, just to the right of the macula. and patient. A good management strategy is easier treatment protocols for diabetic retinopathy if the pa- to implement if the patient and all family caregivers tient ignores them because of the painless, insidious are aware of possible signs and symptoms that are nature and progression of the condition. The common commonly found in diabetic patients. Because diabetic ocular and periocular manifestations of DM are listed retinopathy does not hurt (the only nerve in the eye in Table 8.2. is the sensory optic nerve which has no pain recep- There is great incentive for aggressively treating tors), the patient must be educated about common vi- and managing diabetic retinopathy. Blindness or se- sion changes that are harbingers of diabetic retinopathy vere visual impairment can result from a number and the vision-threatening progression of the disease. of DM-caused conditions including, in order of de- There is no point in having excellent prevention and creasing frequency, vitreous or preretinal haemorrhage, 106 CH 08 MANAGEMENT OF EYE DISEASE AND VISUAL LOSS

Figure 8.3 Pre-proliferative retinopathy. More extensive pathological retinal changes are characteristic of chronic diabetes mellitus. In the early stages of the disease, the haemorrhages are small but can be seen in all quadrants of the image.

Figure 8.4 Proliferative retinopathy of the disk (NVD). With proliferative diabetic retinopathy, some areas of the retina lose their capillary vessels and become non-perfused. This image shows the resulting new blood vessels on the disk. The new blood vessels are tortuous and numerous. macular oedema or related pigmentary change, macu- 8.4 Treatments of diabetic lar or retinal detachment and neovascular glaucoma [25, 26]. Diabetic retinopathy typically begins as a retinopathy series of non-proliferative abnormalities, and then progresses to proliferative diabetic retinopathy. Macular The key to preventing diabetic retinopathy is the strict oedema can develop at any time during the progression control of DM through the use of nutritional ther- of diabetic retinopathy. Macular ischemia, retinal and apy, exercise and drug treatment. This means that vitreous haemorrhage, and retinal detachment are the education of people with DM plays a critical role primary causes of blindness in patients with diabetic in the management and treatment of DM and, there- retinopathy. fore, of diabetic retinopathy. For the past 25 years, the 8.4 TREATMENTS OF DIABETIC RETINOPATHY 107

Figure 8.5 Proliferative retinopathy elsewhere in the retina (NVE). With proliferative diabetic retinopathy, some areas of the retina lose their capillary vessels and become non-perfused. This image shows the resulting, tortuous new blood vessels in other areas of the retina, away from the disk. treatment of diabetic retinopathy has been replete with must have updated nutritional knowledge (preferably ever-improving surgery within the eyeball, using lasers provided by a dietician) in order to support the patient and other surgical procedures (e.g., pars plana vitrec- in adopting a healthy lifestyle [30]. The team members tomy) that remove the scar tissue and debris resulting must be aware of the purpose and importance of the nu- from haemorrhages. Currently, promising new treat- tritional therapy. They must also understand the role of ments using antibodies, growth factors and steroids to dietary fat in controlling diabetes, know the definition decrease the progressive effects of diabetic retinopathy of the glycaemic index and be able to identify foods are being widely investigated. containing carbohydrates/mono-unsaturated fats in order to be able to assess the value and appropriateness 8.4.1 Nutritional treatment of reported diets [31]. Nutritional therapy is an integral component in main- 8.4.2 Exercise therapy taining normal blood glucose levels and in treating the chronic complications of DM. Tobacco use and Physical activity is a powerful tool for helping persons diet – in particular the consumption of fatty acids with DM to achieve metabolic goals [32]. Because and dietary fibre – are significantly associated with intensive glycaemic and blood pressure control has the rate of progression of diabetic retinopathy and been shown to delay both the onset and progression retinopathy-related risk factors [27]. Diets created with of diabetic retinopathy, exercise protocols that help re- the aim of achieving good metabolic control have reduce glycaemia and blood pressure should delay the sulted in the excellent control of diabetic retinopathy progression of diabetic retinopathy. Longitudinal stud- [28]. The American Diabetes Association [29] rec- ies have indicated significant improvements in glucose ommends that such diets be based on individual as- metabolism with aerobic exercise training, and also sessment and treatment goals and outcomes. The di- with resistance training in middle-aged and older men ets should take into consideration usual eating habits and women [33]. Because older adults with type 2 and other lifestyle factors. This personalized approach DM have mobility impairment and reduced fitness, a requires that ongoing nutrition self-management edu- number of investigations are ongoing which aim to cation and care be provided by the health care team optimize exercise programmes, albeit with only lim- for individuals with DM. Although a diet prescribed ited success. For example, a specially developed form by a registered dietician should form the basis for the of Tai Chi, while developed specifically for diabetic nutritional treatment, all members of the medical team patients, may not have been of sufficient intensity, 108 CH 08 MANAGEMENT OF EYE DISEASE AND VISUAL LOSS

Table 8.2 Visual problems resulting from diabetes frequency or duration to effect any positive changes mellitus. in many aspects of physiology or health status relevant to older people with DM [34]. Further research Site or condition Visual system consequences is clearly necessary before specific exercise protocols Refraction Fluctuations in refractive error can be identified for slowing the progression of diabetic Distorted vision retinopathy. Blurred vision Tear films Dry eyes 8.4.3 Surgical treatment Pupil Rubeosis Until very recently, no pharmacological agent has Cornea Ulcers demonstrated the ability to slow the progression of dia- Abrasions betic retinopathy. Since the advent of laser technology Hypoesthesia during the 1980s, pan retinal photocoagulation laser Poor epithelial healing surgery has been used with great success in the treat- Iris Neovascularization ment of macular edema [35] and vitreous haemorrhage Neovascular glaucoma [36, 37] that accompany proliferative diabetic retinopa- Lens Premature cortical cataracts thy. While focal laser photocoagulation reduces the risk Posterior subcapsular cataracts of moderate visual loss by 50–70% in eyes with macu- [23] lar oedema, the downside of these procedures is a lack of efficacy in some patients, patient discomfort, the Ciliary body Premature presbyopia need for repeated treatment, constriction of peripheral Extraocular muscles Cranial nerve palsies visual fields, decreased night vision, reduced near vi- Nerve III sion, loss of acuity, and the risk of retinal damage and Nerve IV (diplopia) Nerve VI (diplopia) scarring. Early vitrectomy improves visual recovery in patients with proliferative retinopathy with accompa- Eyelids Cranial nerve VII palsy (Bell’s nying severe vitreous haemorrhage. The intravitreal palsy) injection of steroids can be used when conventional Vitreous Detachments treatments have failed in eyes, when there is a persis- Optic nerve Ischaemic optic neuropathy tent loss of vision. Nonetheless, laser photocoagulation Neovascularization of the disc and vitrectomy remain the conventional management protocols for diabetic retinopathy [38]. Macula Maculopathy • Oedematous • Exudative 8.4.4 Pharmacological approaches • Ischaemic The ultimate goal of any treatment of diabetic retinopa- Macula oedema thy is the prevention of vision loss. Traditionally, the pharmacological methods to achieve this goal have Retina Detachments Soft exudates been limited to strict metabolic control and tight blood Hard exudates pressure control, and these have been used successfully Retinal oedema to reduce the risk of moderate and severe visual loss by Vein occlusions 50% of patients with severe non-proliferative and pro- Microaneurysms liferative diabetic retinopathy [39]. However, to better Artery occlusions address the 50% of patients who have been unable to Neovascularizations Intraretinal haemorrhages control diabetic retinopathy using these methods, there is an active research agenda to develop new pharma- Trabecular network Increased incidence of open-angle ceuticals that will better protect visual acuity, macular (Ganglion cell glaucoma [24] thickness and, therefore, patient quality of life. apoptosis) 8.5 CONCLUSIONS 109 A growing number of biological agents such as betes control. Better diabetes self-care understanding, cytokines, monoclonal antibodies and fusion proteins self-efficacy and behaviors related to glycaemic con- have become available for the treatment of various trol can be achieved through educational strategies that autoimmune, neoplastic, cardiovascular, infectious, al- provide information to patients, thus motivating them lergic and other conditions. Their introduction has to effectively manage their DM [46, 47]. Educational resulted in marked clinical improvements for many tools such as reminder cards, informational brochures, patients in a variety of conditions. Many new stud- self-care assessments, lists of available diabetes series on the treatment of diabetic retinopathy have been vices and daily recording sheets all enhance patient initiated based on the observations that microvascu- health literacy and compliance. Personal control over lar damage to patients with chronic hyperglycaemia one’s own health care has been shown to improve DM is mediated by interrelated pathways involving aldose control among patients who were taught to self-monitor reductase, advanced glycation end products, protein ki- their blood glucose in France [48], Turkey [49] and the nase C (PKC) and vascular endothelial growth factor United States [50]. Clear and concise patient education (VEGF). Thus, a variety of promising new therapies material that is culturally competent is critical. In addi- for diabetic retinopathy-targeting pathways that cause tion, all interdisciplinary team members providing care microvascular damage are under investigation [39, 40]. to the diabetic patient must give consistent and similar Many of these therapies involve direct growth factor advice and information, which in turn means that there modulators, including VEGF inhibitors, PKC inhibitors must be a coordination of educational materials and of and steroids. Several investigations involving human the message given to the patient (Table 8.3). subjects are currently under way, including a combi- This consistent, culturally sensitive, educational ap- nation of pan retinal photocoagulation and intravitreal proach is particularly important for the elderly pa- injections of triamcinolone acetonide (IVTA) [41, 42] tient with diabetic retinopathy [51], as this condition to treat proliferative diabetic retinopathy; intravitreal is often just one of many ‘geriatric diseases’ that a injections of antibodies against VEGF to treat diabetic patient might have. Other common conditions or dis- oedema [43, 44]; and the use of antibodies against eases might include cataracts, arthritis, cancer, cardiac VEGF, growth factors and steroids to treat neovascu- disease, pulmonary disease, incontinence, cognitive de- larization [45]. All of these studies currently are under cline and sleep deprivation, to mention a few. Some of way, but no specific treatment recommendations have yet been made. Each of the investigating groups has those diseases/conditions have a direct effect on the emphasized the need for a more comprehensive eval- health of the eye, and must be considered when de- uation in a multicenter randomized controlled clinical veloping treatment plans for diabetic retinopathy (see trial with longer follow-up before clinical guidelines Table 8.3). Today, the management and treatment of can be developed. a single disease is no longer sufficient for elderly patients, who tend to become a ‘treatment unit’ rather than a particular disease. That said, the active manage- 8.5 Conclusions ment of diabetic retinopathy does need to be addressed because it has the potential to progress to visual disabil- Relatively recent improvements in the understanding ity and/or blindness. Visual loss has a major impact on of a need for tight glycaemic and blood pressure quality of life of an older person because it contributes control have focused the treatment of DM into four not only to morbidity but also to mortality [52, 19, 4, main strategies, namely exercise, diet, surgery and 53]. If the patient wishes to actively address the DM pharmaceuticals. Each of these strategies plays an then the health care team should provide the educa- important role in reducing the incidence of blindness tional materials that will assist in that goal. Yet, if the caused by diabetic retinopathy. However, this broad loss of sight is not a priority, then the team must also front of treatment options relies on excellent patient respect that decision. The interdisciplinary health care education for its success. team, when developing educational interventions, must Patient knowledge of actual and target HbA1c val- avoid two pitfalls that are particularly apropos to the ues is needed for effective patient involvement in dia- geriatric diabetic patient: 110 CH 08 MANAGEMENT OF EYE DISEASE AND VISUAL LOSS

Table 8.3 The importance of maintaining healthy eyes. Medical conditions or diseases Effect on healthy eyes Poor glycaemic control Fluctuation in visual acuity Distorted vision Cataracts High blood pressure Retinal arteriolar narrowing Blurred vision Glaucoma Cardiovascular disease/stroke Central retinal artery occlusions Central retinal vein occlusions Peripheral visual ﬁeld loss Glaucoma Peripheral vascular disease Anterior ischemic optic neuropathy Central retinal artery occlusions Central retinal vein occlusions Retinal haemorrhaging Glaucoma Atherosclerosis Increased frequency and extent of retinal haemorrhages Increased microaneurysms and ischaemic areas Retinal venous occlusions Macular degeneration Poor blood ﬂow Retinal oedema

• The team must not project an attitude of fear of vi- 3. Wild S, Roglic G, Green A, Sicree R and King H. sion loss, nor of defeatism, which might indicate (2004) Global prevalence of diabetes: estimates for the a lack of faith in the patient’s ability to manage year 2000 and projections for 2030. Diabetes Care 27 diabetic retinopathy along with other coexisting con- (5): 1047–53. ditions and diseases. 4. Morello CM. (2007) Etiology and natural history of diabetic retinopathy: an overview. Am J Health Syst • The team must also provide support for the patient’s Pharm, 64 (17 Suppl. 12): S3–7. health care priorities, which may or may not rate the 5. Healthy People 2020. http://www.healthypeople.gov/ prevention of diabetic retinopathy as highly as does (last accessed 1 October 2007). the team. In other words, the team must not practice 6. Cikamatana L, Mitchell P, Rochtchina E, Foran S and excessive interventionism that might be inappropri- Wang JJ. (2007) Five-year incidence and progression ate to the patient’s wishes. of diabetic retinopathy in a deﬁned older population: Only through education and communication can the the Blue Mountains Eye Study. Eye 21 (4): 465–71. entire health care team – which includes not only 7. Roy, MS. (2000) Diabetic retinopathy in African Amer- the patient but also the family and the healthcare icans with type 1 diabetes: The New Jersey 725: providers – make the appropriate decisions to treat the I. Methodology, population, frequency of retinopa- entire patient, and not just a single disease. thy, and visual impairment. Arch Ophthalmol 118 (1): 97–104. 8. Harris MI, Hadden WC, Knowler WC and Bennett PH. References (1987) Prevalence of diabetes and impaired glucose tolerance and plasma glucose levels in U.S. population 1. NIH www.nei.nih.gov/eyedata (last accessed 13 Octo- aged 20-74 yr. Diabetes 36 (4): 524–34. ber 2007). 9. Roy MS and Affouf M. (2006) Six-year progression 2. World Health Organization. http://www.who.int/ of retinopathy and associated risk factors in African mediacentre/factsheets/fs282/en/ (last accessed 25 American patients with type 1 diabetes mellitus: the November 2007). New Jersey 725. Arch Ophthalmol, 124 (9): 1297–306. 8.5 CONCLUSIONS 111

10.VarmaR,TorresM,PenaF,KleinRandAzenSP. 23. Klein BE, Klein R and Lee KE. (1998) Diabetes, (2004) Prevalence of diabetic retinopathy in adult Lati- cardiovascular disease, selected cardiovascular disease nos: the Los Angeles Latino eye study. Ophthalmology risk factors, and the 5-year incidence of age-related 111 (7): 1298–306. cataract and progression of lens opacities: the Beaver 11. Bloomgarden ZT. (2007) Screening for and managing Dam Eye Study. Am J Ophthalmol 126 (6): 782–90. diabetic retinopathy: current approaches. Am J Health 24. Nakamura M, Kanamori A and Negi A. (2005) Dia- Syst Pharm. 64 (12 Suppl. 12): S8–14. betes mellitus as a risk factor for glaucomatous optic 12. Klein R, Klein BE, Moss SE and Cruickshanks KJ. neuropathy. Ophthalmologica 219 (1): 1–10. (1998) The Wisconsin Epidemiologic Study of Diabetic 25. Fong DS, Ferris FL, III, Davis MD and Chew EY. Retinopathy: XVII. The 14-year incidence and progres- (1999) Causes of severe visual loss in the early treat- sion of diabetic retinopathy and associated risk factors ment diabetic retinopathy study: ETDRS Report no. 24. in type I diabetes. Ophthalmology 105 (10): 1801–15. Early Treatment Diabetic Retinopathy Study Research 13. Anderson D, Horton C and O’Toole M. (2007) Integrat- Group. Am J Ophthalmol 127 (2): 137–41. ing depression care with diabetes care in real-world set- 26. Fong DS, Sharza M, Chen W, Paschal JF, Ariyasu RG tings: Lessons from the Robert Wood Johnson Founda- and Lee PP. (2002) Vision loss among diabetics in a tion Diabetes Initiative. Diabetes Spectrum 20: 10–16. group model health maintenance organization (HMO). 14. Roy MS, Roy A and Affouf M. (2007) Depression is Am J Ophthalmol 133 (2): 236–41. a risk factor for poor glycemic control and retinopathy 27. Cundiff DK and Nigg CR. (2005) Diet and diabetic in African-Americans with type 1 diabetes. Psychosom retinopathy: insights from the Diabetes Control and Med 69 (6): 537–42. Complications Trial (DCCT). MedGenMed 7(1):3. 15. Rodin G, Olmsted MP, Rydall AC, Maharaj SI, Colton 28. Hansson-Lundblad C, Agardh E and Agardh CD. PA, Jones JM, Biancucci LA and Daneman D. (2002) (1997) Retinal examination internals in diabetic patients on diet treatment only. Acta Ophthalmol Scand Eating disorders in young women with type 1 diabetes 75 (3): 244–8. mellitus. J Psychosom Res 53 (4): 943–9. 29. The American Diabetes Association. (2004) Nutrition 16. Steinhart JR, Sugarman JR and Connell FA. (1997) recommendations and principles for people with dia- Gestational diabetes is a herald of NIDDM in Navajo betes mellitus. Diabetes Care 23 (Suppl. 2): S43–6. women. High rate of abnormal glucose tolerance after 30. Koura MR, Khairy AE, Abdel-Aal NM, Mohamed HF, GDM. Diabetes Care 20 (6): 943–7. Amin GA and Sabra AY. (2001) The role of primary 17. Mitchell P, Smith W, Wang JJ and Attebo K. (1998) health care in patient education for diabetes control. Prevalence of diabetic retinopathy in an older commu- Egypt Public Health Assoc 76 (3-4): 241–64. nity. The Blue Mountain Study. Ophthalmology 105 31. Heller T, Maisios M and Shahar D. (2007) Physicians’ (3): 406–11. and nurses’ knowledge and attitude towards nutritional 18. Juutilainen A, Lehto S, Ronnemaa T, Pyorala K and therapy in diabetes. Harefuah, 146 (9): 670–4. Laakso M. (2007) Retinopathy predicts cardiovascular 32. Franz MJ. (1997) Lifestyle modifications for diabetes mortality in type 2 diabetic men and women. Diabetes management. Endocrinol Metab Clin North Am 26 (3): Care 30 (2): 292–9. 499–510. 19. Lovestam-Adrian M, Hansson-Lundblad C and Torffvit 33. Ryan, AS. (2000) Insulin resistance with aging: effects O. (2007) Sight-threatening retinopathy is associated of diet and exercise. Sports Med 30 (5): 327–46. with lower mortality in type 2 diabetic subjects: a 34. Tsang T, Orr R, Lam P, Comino EJ and Singh MF. 10-year observation study. Diabetes Res Clin Pract.77 (2007) Health benefits of Tai Chi for older patients with (1): 141–7. type 2 diabetes: the “Move It For Diabetes study” – a 20. Suto C, Hori S, Kato S, Muraoka K and Kitano S. randomized controlled trial. Clin Interv Aging 2(3): (2006) Effect of perioperative glycemic control in pro- 429–39. gression of diabetic retinopathy and maculopathy. Arch 35. Fong DS, Strauber SF, Aiello LP, Beck RW, Callanan Ophthalmol 124 (1): 38–45. DG, Danis RP, David MD, Feman SS, Ferris F, Fried- 21. The American Diabetes Association’s Clinical manSM,GarciaCA,GlassmanAR,HanDP,LeD, Practice Guidelines (http://diabetes.org/for-health- Kollman C, Lauer AK, Recchia FM and Solomon SD. professionals-and-scientists/cpr.jsp (last accessed 27 (2007) Writing Committee for the Diabetic Retinopa- October 2007). thy Clinical Research Network. Arch Ophthalmol 125 22. Onofrey BE, Skorin L, Jr and Holdeman NR. (eds) (4): 469–80. (2005) Ocular Therapeutics Handbook - A Clinical 36. Lovestam-Adrian M, Agardh CD, Torffvit O and Manual, 2nd edition. Lippincott Williams and Wilkins, Agardh E. (2003) Type 1 diabetes patients with Philadelphia. severe non-proliferative retinopathy may benefit from 112 CH 08 MANAGEMENT OF EYE DISEASE AND VISUAL LOSS

panretinal photocoagulation. Acta Ophthalmol Scand 46. Heisler M, Smith DM, Hayward RA, Krein SL and 81 (3): 221–5. Kerr EA. (2003) How well do patients’ assessments 37. Luttrull JK, Musch DC and Spink CA. (2007) Sub- of their diabetes self-management correlate with actual threshold diode micropulse panretinal photocoagula- glycemic control and receipt of recommended diabetes tion for proliferative diabetic retinopathy. Eye,22(5), services? Diabetes Care 26 (3): 738–43. 607–12. 47. Heisler M, Piette JD, Spencer M, Kieffer E and Vijan S. 38. Mohamed Q, Gillies MC and Wong TY. (2007) Man- (2005) The relationship between knowledge of recent agement of diabetic retinopathy: a systematic review. HbA1c values and diabetes care understanding and JAMA 298 (8): 902–16. self-management. Diabetes Care 28 (4): 816–22. 39. Yam JC and Kwok AK. (2007) Update on the treatment 48. Guerci B, Drouin P, Grange V, Bougneres P, Fontaine of diabetic retinopathy. Hong Kong Med J 13 (1): P, Kerlan V, Passa P, Thivolet CH, Vialettes B 46–60. and Charbonnel B, for the ASIA Group. (2003) 40. Ryan GJ. (2007) New pharmacologic approaches to Self-monitoring of blood glucose signiﬁcantly treating diabetic retinopathy. Am J Health Syst Pharm improves metabolic control in patients with type 2 64 (17 Suppl. 12): S15–21. diabetes mellitus: the Auto-Surveillance Intervention 41. Choi KS, Chung JK and Lin SH. (2007) Laser pho- Active (ASIA) study. Diabetes Metabol 29 (6): tocoagulation combined with intravitreal triamcinolone 587–94. acetonide injection in proliferative diabetic retinopa- 49. Ozmen B and Boyvada S. (2003) The relationship be- thy with macular edema. Korean J Ophthalmol 21 (1): tween self-monitoring of blood glucose control and 11–17. glycosylated haemoglobin in patients with type 2 dia- 42. Zein WM, Noureddin BN, Jurdi FA, Schakal A and betes with and without diabetic retinopathy. J Diabetes Bashshur ZF. (2006) Panretinal photocoagulation and Complications 17 (3): 128–34. intravitreal triamcinolone acetonide for the manage- 50. Murata GH, Shah JH, Hoffman RM, Wendel CS, Adam ment of proliferative diabetic retinopathy with macular KD, Solvas PA, Bokhari SU and Duckworth WC. edema. Retina 26 (2): 137–42. (2003) Intensiﬁed blood glucose monitoring improves 43. Arevalo JF, Fromow-Guerra J, Quiroz-Mercado glycemic control in stable, insulin-treated veterans with H, Sanchez JG, Wu L, Maia M, Berrocal MH, type 2 diabetes: the Diabetes Outcomes in Veterans Solis-Vivanco A, Farah ME and the Pan-American Study (DOVES). Diabetes Care 26 (6): 1759–63. Collaborative Retina Study Group. (2007) Primary 51. Massin P and Kaloustian E. (2007) The elderly dia- intravitreal bevacizumab (Avastin) for diabetic betic’s eyes. Diabetes Metab 33 (Suppl. 1): S4–9. macular edema: results from the Pan-American 52. Hirari FE, Moss SE, Klein BE and Klein R. (2008) Collaborative Retina Study Group at 6-month Relationship of glycemic control, exogenous insulin, follow-up. Ophthalmology 114 (4): 743–50. and C-peptide levels to ischemic heart disease mortal- 44. Haritoglou C, Kook D, Neubauer A, Wold A, Prigliner ity over a 16-year period in persons with older-onset S, Strauss R, Gandorfer A, Ulbig M and Kampik A. diabetes: The Wisconsin Epidemiologic Study of Di- (2006) Intravitreal bevacizumab (Avastin) therapy for abetic Retinopathy (WESDR). Diabetes Care,31(3), persistent diffuse diabetic macular edema. Retina 26 493–7. (9): 999–1005. 53. West SK, Munoz B, Istre J, Rubin GS, Friedman SM, 45. Emerson MV and Lauer AK. (2007) Emerging thera- Fried LP, Bandeen-Roche K and Schein OD. (2000) pies for the treatment of neovascular age-related macu- Mixed lens opacities and subsequent mortality. Arch lar degeneration and diabetic macular edema. BioDrugs Ophthalmol 118 (3): 393–7. 21 (4): 245–57. 9 The Diabetic Foot

Andrew J. M. Boulton1 and Matthew J. Young2 1Manchester Royal Inﬁrmary, Manchester, UK 2Edinburgh Royal Inﬁrmary, Edinburgh, UK

Excellence (NICE) has published guidelines of Key messages foot care for patients with type 2 diabetes [4]. However, evidence is at last beginning to appear • All patients with diabetes should be screened that structured diabetic foot care, when performed annually for risk of foot ulceration; those with in a multidisciplinary team, does eventually result in significant risk factors require more frequent significant improvements in amputation rates among review and education in self-foot care. people with diabetes [5]. • Neuropathic foot ulcers under pressure areas Foot problems in diabetes can develop from a will heal if the ulcer is adequately off-loaded; number of component causes. The main contribut- this is frequently neglected in clinical practice. • Most diabetic foot ulcers will heal if the arterial ing factors include sensorimotor and autonomic neu- inflow is sufficient, any infection is treated ropathy, peripheral vascular disease (PVD), limited aggressively, and the pressure is removed from joint mobility and high foot pressures. The existence the ulcer area. of other long-term complications of diabetes (partic- • Any patient presenting with a unilateral, warm ularly end-stage renal disease as a consequence of swollen foot in the presence of neuropathy nephropathy) also influence the development of foot should be assumed to have an acute Charcot ulceration. Clearly, general practitioners, geriatricians joint, until proven otherwise. and diabetologists must all pay particular attention to the feet of older patients to prevent significant avoidable morbidity and mortality in this vulnerable 9.1 Introduction group. Lower-limb amputation is more common in older, The St Vincent timescale [1] to reduce the number of usually type 2, diabetic patients [6]; indeed, the average age of diabetic foot clinic attendees is over 60 years, amputations for diabetes in Europe by 50% within 3 which clearly shows that the elderly are at particular years seems a dim and distant memory to those who risk of foot ulceration. Reduced mobility (particularly work with the diabetic foot. The publication of BDA at the hip) in patients aged over 60 impairs their ability = ( RCP) guidelines in England and Wales, and SIGN to inspect the feet, and leads to the continued progres- guidelines in Scotland, have demonstrated that there sion of foot lesions that are often beyond the point of is still a lack of good evidence-based randomized repair even before they are discovered [7]. Very of- controlled trials on which to base diabetic foot care ten, patients with severely impaired vision depend on [2, 3]. More recently, the National Institute of Clinical other people to inspect their feet, but when this is not

Diabetes in Old Age Third Edition Edited by Alan J. Sinclair © 2009 John Wiley & Sons Ltd. ISBN: 978-0-470-06562-4 114 CH 09 THE DIABETIC FOOT possible it may be very difficult for the patient to ensure unmyelinated) fibre involvement, a loss of vibration that their foot care is adequate. perception and proprioception is believed to be related Diabetes alone probably does not add to the to large (usually myelinated) fibre damage. Painful prevalence of bunions, clawed toes and medial arterial symptoms are found in approximately 11% of all calcification that is seen in the elderly [8, 9]. Neuropa- diabetic patients, and can be particularly distressing. thy, however, is more prevalent in the elderly, and The therapies for painful diabetic neuropathy vary increases with both age and the duration of diabetes. from non-pharmacological interventions such as tran- Once this is superimposed on the normal aging scutaneous nerve stimulation or complementary ther- process, then skeletal abnormalities (including sponta- apies to drugs with potentially major side effects. In neous fractures) become significantly more common each case there is often a clear placebo effect, and [9]. If the increased prevalence of PVD in older type 2 it is often better to start with low doses and build diabetic patients is added to the increase in neuropathy up to an effective dose of any drug therapy. None of then, together with difficulties in personal foot care, the non-analgesic pharmacological agents, tricyclics or this fully explains the particular predilection for foot anti-arrhythmics is currently licensed for this use, and problems that exists in older diabetic patients. The the use of such agents should therefore be explained to demographic changes, increasing numbers of elderly the patient, in detail, including their often significant people, the increasing proportion of those who side-effect profiles, prior to their use. Particular care live alone, and increasing levels of obesity – all with these agents is needed in the elderly population, of which are currently occurring within the United in view of their frequent adverse side effects [13]. The Kingdom – will only serve to add to the already most commonly used drugs are tricyclic antidepres- substantial numbers of type 2 diabetic patients who sants and anti-epileptics, such as gabapentin and pre- develop foot ulceration and peripheral ischaemia. The gabalin. Both, gabapentin and pregabalin are licensed challenge, therefore, is to reduce this excess burden for the treatment of painful neuropathy on the basis of of risk to a minimum by accurate detection and the promising clinical trials: similarly, the dual reuptake amelioration of risk. inhibitor duloxetine, also has demonstrated proven efficacy in large, randomized controlled trials. Although 9.2 Peripheral sensorimotor such therapies are useful, like all adjuvant analgesics they require careful titration in order to minimize the neuropathy side effects [11, 13]. Capsaicin 0.075% cream has been licensed for the treatment of post-herpetic neuralgia for Peripheral sensorimotor neuropathy is a major con- many years, and more recently was licensed for the tributory cause in 80% of diabetic foot ulceration treatment of painful peripheral neuropathy. In addition [10]. The incidence of diabetic peripheral sensorimo- to monotherapy, it can also be safely added to any ex- tor neuropathy increases with the duration of diabetes; isting, partially effective oral agent. Unfortunately, its however, as the prevalence depends on the diagnostic usefulness is often limited by a poor understanding of criteria used, the prevalence rates reported from dif- the need to use it little and often, and to persevere be- ferent epidemiological studies will vary considerably yond the initial first week or two, during which time it [11]. A large UK multicentre study which screened a may even make the symptoms a little worse [14]. large hospital-treated diabetic population found that the It must be remembered that in some patients the overall prevalence of neuropathy was 28.5% [12], with presenting feature of peripheral neuropathy, and more than half of all patients with type 2 diabetes, and indeed of diabetes itself, may be foot ulceration, aged over 60 years, being found to have neuropathy. as the progression to an insensitive foot may occur Therefore, the majority of the elderly population with without any positive symptoms. It is not uncommon diabetes is at an increased risk of foot ulceration. for neuropathic diabetic patients to present because The most common symptoms of sensory neuropathy of the smell caused by purulent discharge. Thus, the are numbness, lancinating pain, ‘pins and needles’, absence of symptoms must never be equated with an burning pain and hyperaesthesiae, typically with absence of risk of ulceration. nocturnal exacerbation. The clinical signs are usually The diagnosis of diabetic peripheral neuropathy sensory loss in a glove-and-stocking distribution for clinical purposes is a complex issue [11, 15]. (see Figure 9.1). Whilst loss of pain, fine touch and The diagnosis of the ‘at-risk’ neuropathic foot during temperature sensation are related to small (often routine screening at the annual review requires 9.2 PERIPHERAL SENSORIMOTOR NEUROPATHY 115

Figure 9.1 Extensive burns on the leg of a neuropathic diabetic patient who had fallen asleep in front of the gas fire, to be woken by the smell of burning. a less-sophisticated clinical examination than the predictive of subsequent foot ulceration [18]. In differential diagnosis of diabetic peripheral neuropathy the latter study, patients with a VPT >25 V were in secondary care, which often demands a more sevenfold more likely to develop a foot ulcer than detailed investigation. For routine screening purposes, a patient with a VPT <25 V over a 4-year period. clinical examination using 10 g monofilaments and, This increased to 11-fold when recurrent ulceration for example, a 128 Hz tuning fork will suffice. The was considered, as no patient with a VPT <25 V requirements for the comprehensive diabetic foot developed a second ulcer. This study also revealed examination to identify the high-risk foot were a relationship between foot ulceration and increasing the subject of a recent task force of the American age although, even after correcting for this, the VPT Diabetes Association [16]. The vibration perception remained a strong predictor of foot ulceration risk. threshold (VPT) can be measured quantitatively Monofilaments represent a quick method for assess- using a Neurothesiometer (Arnold Horwell, UK). ing the at-risk foot in diabetes, and currently are used Such measurements have shown to be increased extensively in the United Kingdom. There is some in association with other measurements of diabetic doubt as to the reliability of manufacture of monofila- peripheral neuropathy, but also increase with normal ments, and widely varying rates of ulcer incidence have aging; therefore the use of age-related normal been reported in studies that have used monofilaments values have been recommended. Other authors have as a screening tool [19]. In general a 10 g monofila- claimed that the increased coefficient of variation in ment should be used in a variety of sites on the foot, older patients makes the measurement of vibration with a clearly defined pass/fail criterion. Monofilament perception threshold unreliable, and that it should remain relatively cheap to buy and easy to use, and be supplemented by other tests of neuropathy [17]. therefore are very popular with many foot clinics [16]. For screening purposes, however, this may not Diabetologists. however, still tend to diagnose be necessary. A VPT >25 V has been shown in peripheral neuropathy on clinical grounds, and this cross-sectional and prospective studies to be strongly approach can be improved by use of a neuropathy 116 CH 09 THE DIABETIC FOOT disability score (NDS) [12]. The NDS is derived from an examination of the ankle reflex vibration sensation, using a 128 Hz tuning fork, a pin-prick sensation and temperature (with a cold tuning fork) sensation at the great toe. Each sensory modality is scored as either normal = 0 or reduced = absent = 1 for each side, and the ankle reflexes as normal = 0, present with reinforcement = 1orabsent= 2 per side. Thus, while the total maximum abnormal score is 10, a score ≥ 6 can be regarded as indicative of significant peripheral neuropathy. Such a score correlates well with the VPT measurements which, as described above, predict foot ulceration in diabetic patients. The NDS was confirmed as being the best predictor for the risk of foot ulceration in a large prospective community study conducted in the UK [20]. If a Neurothesiometer is available, then a VPT >25 V in both feet will predict that up to 84% of such patients will develop foot ulceration over the next four years [18]. Motor fibre loss is another significant result of peripheral neuropathy leading to small muscle atrophy in the foot. As a consequence, there is an imbalance between flexor and extensor muscle function that results in clawing of the toes, prominent metatarsal heads and anterior displacement of the metatarsal footpads (Figure 9.2). Abnormally high foot pressures usually develop under these areas and, as discussed below, can Figure 9.2 At-risk foot, showing prominent metatarsal lead to foot ulceration in the susceptible foot. heads and clawed toes. In addition, the gait pattern is significantly altered in patients with diabetic neuropathy [21]; this may alter the foot pressure distribution and make the foot more twice a day, is required to keep the skin supple and prone to the effects of high pressure. Gait problems, reduce the risk of such fissures. with increasing falls, and the risks of injury to the feet, A loss of sympathetic tone in small vessels also are increased in neuropathic diabetic patients. Such leads to a reduced resistance and increased arteriove- problems are more pronounced in elderly neuropathic nous shunting. The venous P O2 and pressure is raised patients, and even worse in those with visual handicap, in the neuropathic limb to a level approaching that of increasing the risk of foot ulceration. arterial blood, and has been measured at higher levels than in the endoneurium [23]. Thus, in a diabetic patient with autonomic neuropathy, but without coex- 9.3 Autonomic neuropathy isting vascular disease, the blood flow is increased at rest and the distended dorsal foot veins can be seen. Autonomic neuropathy results in a wide spectrum of Initially, the overall increase in blood flow increases problems in the cardiovascular, gastrointestinal and capillary pressure; however, over time this leads to mi- genitourinary systems, and is known to be associated crovascular sclerosis and, when taken in conjunction with the development of foot ulceration in diabetic with the increased shunting in the diabetic neuropathic patients [22]. In the foot, denervation of the sweat foot, it may lead to inadequate nutritional flow and glands leads to dry, atrophic skin and callus formation. subsequent tissue ischaemia, greatly increasing the risk Severe cracking of the skin often occurs under these of ulceration [24]. The coexistence of autonomic neu- circumstances, and facilitates microbial infections. The ropathy and macrovascular disease may cause a further regular use of emollient creams or ointments, often deterioration in the level of tissue oxygenation. 9.4 PERIPHERAL VASCULAR DISEASE 117 9.4 Peripheral vascular disease to the increase in medial arterial calcification associated with diabetic nephropathy, which is an indepen- Both, the micro- and macrocirculation in the lower ex- dent marker of increased mortality in diabetes [29]. tremities are affected by diabetes. In the microcircula- Diabetes alone does not increase the prevalence of tion, the skin capillary pressure is increased in patients medial arterial calcification in matched groups of con- with type 1 diabetes (of either recent onset or of long trols and non-neuropathic subjects, although there is duration), and this abnormality reverses when the di- significantly heavier arterial calcification in the feet of abetes control is improved [25]. This increase in the neuropathic diabetic patients [9]. The VPT, duration capillary pressure is most likely responsible for the loss of diabetes and serum creatinine are all independent of the blood flow autoregulation, increased arteriove- predictors of the degree of medial arterial calcification which, when present, is known to alter the pulse wave- nous shunting, impaired hyperaemic response, changes form and falsely elevate ankle pressures in diabetic in capillary blood flow and basement membrane thick- patients [30]. Therefore, it has been suggested that ening seen in diabetic patients. This microvascular toe systolic pressure measurements might replace ankle sclerosis may contribute to nephropathy, retinopathy pressure measurements as an index of arterial inflow and probably also neuropathy, but the direct role in to the diabetic foot, as the ankle pressure index (mea- the development of foot ulceration remains unclear. sured with Doppler ultrasound) may be misleadingly Macrovascular disease is more common in diabetic high despite the presence of occlusive PVD. If the an- patients. Peripheral vascular disease is estimated to kle systolic pressure is more than 75 mmHg above the occur at least twice as commonly in diabetic than brachial systolic pressure, then this is highly indicative in non-diabetic patients [26]. Lipid disorders, platelet of medial arterial calcification and the poor reliabil- dysfunction, increased coagulation and endothelial cell ity of the ankle pressure to indicate lower extremity dysfunction have been implicated in the pathogene- arterial disease [31]. Since it is also possible that me- sis of the atherosclerosis. Peripheral vascular disease dial arterial calcification will falsely elevate the ankle usually has the same clinical presentation as that seen pressure into the normal range in some neuropathic in non-diabetic patients, with intermittent claudication, patients, a normal ankle pressure should be interpreted rest pain, ulceration and gangrene being the main clin- with caution, and perhaps be complemented with plain ical features [26]. However, the symptoms may be radiography of the foot, particularly in elderly and masked by coexisting peripheral neuropathy, and sig- neuropathic diabetic patients. A low ankle pressure, nificant ischaemia may develop in the absence of pain. together with an ankle pressure index <0.9, suggests Although the femoropopliteal segments are most of- the presence of arterial occlusive disease and the need ten affected, small vessels below the knee, such as the for further investigation [31]. However, despite all of tibial and peroneal arteries, are more severely affected the problems associated with ankle pressure measure- in diabetic than in non-diabetic patients [27]. This ments, foot pulses remain the best clinical guide to the means, overall, that vascular disease in diabetic pa- presence of PVD in diabetes. Nonetheless, the Amer- tients is more likely to lead to amputation even though, ican Diabetes Association has recommended that all level for level, the outcome of revascularization is sim- diabetic patients aged >50 years should undergo an ilar to that in the non-diabetic population. In addition, annual ankle brachial index check [32]. Diabetic pa- the presence of simultaneous cardiac and cerebrovas- tients with neuropathy may have significant ischaemia cular disease means that the long-term survival after with no pain because of the loss of pain sensation. such procedures is often shorter. Medial arterial calci- The absence of foot pulses indicates the presence of fication is another common finding in diabetic patients, vascular disease, even if the popliteal pulse is present and can be recognized on X-ray films by its ‘pipe-stem’ and there is no complaint of claudication or rest pain. appearance. Medial arterial calcification is reported to Any areas of cyanosis or peripheral necrosis are also be associated with diabetic peripheral somatosensory indicative of arterial insufficiency. and autonomic neuropathy. Diabetic patients with evidence of PVD should be Previous studies of the quantitative distribution referred for vascular assessment with arterial recon- of medial arterial calcification within the diabetic struction or angioplasty where appropriate [26]. The foot have shown the condition to be significantly best advice for any stable claudicant with no evidence associated with an increased prevalence of cardiovas- of tissue loss is to ‘stop smoking and keep walking’. cular mortality [28], although this may also be related Indeed, there is considerable evidence that, in order 118 CH 09 THE DIABETIC FOOT to be effective, the patient should walk to the point of claudicating, and even for some distance with claudication. It is believed that this might encourage the proliferation of collateral circulation [33]. Although the long-term benefits of reconstructive surgery for diabetic patients with claudication remain the subject of dispute among vascular surgeons, this approach should be considered if the walking distance is reduced, even before the onset of tissue loss. A reducing claudication distance is a sign of impending critical ischaemia in diabetic patients, who have a greater tendency towards an early and more aggressive progression of arterial disease. Most vascular teams will not consider patients with stable claudication for surgery, and will generally wait for the development of critical limb ischaemia prior to intervention. However, diabetic patients also have a higher amputation rate than non-diabetic patients for similar initial grades of arterial disease [34]. With such clear evidence that tissue loss has a significantly adverse effect on limb prognosis, surgery before the onset of critical ischaemia has its advocates in many centres, and should be encouraged. As an additional consideration, peripheral autonomic neuropathy is usually present in patients with foot ulceration; the patients can therefore be said to have performed auto-sympathectomy. Surgical sympathectomy Figure 9.3 The ‘prayer sign’. is still occasionally attempted in a number of diabetic patients but, because of the pre-existing peripheral au- shock-absorbing capacity during walking [39]. This re- tonomic changes and associated medial arterial calcifi- sults in increased plantar foot pressures and, in the cation, it is unlikely to produce any substantial benefit. neuropathic foot, may be a contributory factor to the development of foot ulceration [40]. Limited extension of the great toe, known as ‘hallux rigidus’, can 9.5 Limited joint mobility also predispose to ulceration by limiting the adaptive extension of the toe during the final ‘toe-off’ phase of Diffuse collagen abnormalities are common in diabetic walking, thus increasing the vertical and shear forces patients [35]. The main pathogenic mechanism for on the toe. these abnormalities is a glycation of collagen, which results in the thickening and increased crosslinking of collagen bundles [36]. One of the clinical mani- 9.6 Foot pressure abnormalities festations of this change is a thick, tight and waxy skin, leading to a restriction of joint movements. The two main factors responsible for the development Patients with such limited joint mobility are unable of high foot pressures, motor neuropathy and limited to oppose the palms of their hands (the ‘prayer joint mobility, have already been discussed. Callus sign’; see Figure 9.3). The term ‘cheiroarthropathy’ formation, which itself is a result of high foot pressures has been used to describe this condition, although as and dry skin, may also act as a foreign body and result other joints – including those in the shoulder, hip and in further increases of these pressures [41]. In contrast, foot – can also be affected a more appropriate term the patient’s age and bodyweight do not significantly is ‘limited joint mobility’, and this is now in general influence the foot pressure, most likely because the use [37, 38]. Limited joint mobility in the foot mainly foot contact surface area also increases with weight involves the subtalar joint, which provides the foot with [42, 43]. 9.7 OTHER RISK FACTORS 119 Intermittent moderate stress on healthy tissue for an Studies of patients with traumatic or diabetic ampu- excessive time – as in the case of abnormal pressures tations have shown that amputation alone does not applied on the plantar surface of the foot during ex- cause an increase in the loads under the remaining foot. cessive walking – can lead to tissue inflammation and, However, in a neuropathic diabetic patient amputation finally, to ulceration. At the microscopic level, it is is associated with high pressure under the remaining believed that pressure overcomes the nutritive capillary foot, probably related to neuropathy and limited joint blood flow of the skin, and this leads to a localized mobility in that foot [50]. tissue necrosis and breakdown [44]. The demonstra- In a prospective study of the prediction of foot tion of increased arteriovenous shunting in the diabetic ulceration using VPTs, recurrent ulceration was com- foot, and the reports of an impaired hyperaemic injury mon, affecting over 50% of the patients. In addition, response in neuropathic patients [45, 46] may also in an audit of dressing policy at the Manchester Foot contribute to the increased risk of ulceration. Studies in Hospital, the median number of ulcers per patient was dogs have shown that repetitive moderate trauma leads two (range 1 to 12) [51]), again suggesting that over to the eventual breakdown of the skin and ulceration half the patients re-ulcerated despite preventive care [44]. Thermography of the feet of patients with diabetic and advice. The likely causes for this are unknown, but neuropathy has shown hot spots of inflammation in ar- it appears that those patients who do not wear their rec- eas of high foot pressures and repetitive trauma [47]. ommended shoes (which are usually supplied), or do Sensory dysfunction is crucial for the development of not follow the appropriate advice, are those who sub- neuropathic ulceration. In a non-neuropathic subject, sequently re-ulcerate. Strategies to increase footwear the pain which accompanies the inflammation will usu- acceptability and compliance can reduce re-ulceration ally force the individual to rest the foot before it pro- rates. gresses to ulceration, whereas a patient with a loss of pain awareness will continue to walk long after an ulcer has developed. Therefore, high foot pressures alone, in 9.7.2 Reduced resistance to infection the absence of sensory neuropathy, do not result in There are many reasons for impaired resistance to in- foot ulceration. This can be illustrated in patients with fection in a diabetic ulcer. Diabetes is associated with rheumatoid arthritis, in whom joint involvement in the impaired neutrophil function, particularly in the pres- feet results in high foot pressures comparable to those ence of a high blood glucose, and both macrocircula- found in diabetic patients, but not in ulceration [48]. tory and microcirculatory abnormalities lead to relative The measurement of foot pressures is not routine in hypoxia in the wound [52]. Multiple microbes (often a most clinics. A careful visual inspection, with palpa- mixture of aerobic and anaerobic bacteria) are usually tion of the foot, can detect most high-pressure areas, found in cultures from foot ulcers. The most com- and accommodative insoles can therefore be made to mon pathogenic organisms in diabetic foot ulcers are redistribute pressure away from vulnerable areas, with- Staphylococcus and Streptococcus sp.; the streptococci out the need for expensive foot pressure measuring are often faecal in origin. The clinical relevance of or- systems. A simple foot pressure map, the Podotrack ganisms grown from superficial swabs is variable, as or PressureStat, has been shown to be accurate when other organisms may colonize the wound surface and compared to complex optical foot pressure systems, the quality of the sample and method of transport and and may also be used in patient education as the culture may markedly influence the reliability of the high-pressure areas are a dark grey or black colour, result [53]. The treatment of infections associated with which emphasize to the patient which particular areas foot ulceration is detailed further in the management under the foot are at greatest risk [49]. of foot ulceration (see below).

9.7 Other risk factors 9.7.3 Smoking and alcohol Smoking is known to be associated with foot ulcera- 9.7.1 History of previous foot problems tion, probably by increasing the prevalence of vascular A history of previous foot problems in a diabetic disease [54]. Recurrent neuropathic foot ulceration has patient strongly suggests that they are at high risk been reported as being more common in patients with for future problems, especially lower-limb amputation. high alcohol consumption [55]. 120 CH 09 THE DIABETIC FOOT

9.7.4 Other complications of diabetes Wagner, although both were good predictors of outcome [59]. Foot pressures in patients with nephropathy are higher There are, however, a number of classification sys- than in diabetic patients without renal impairment. tems currently in use at different centres, and with such In combination with neuropathy, which is also more a variety available it is clear that no one scheme can of- common in such patients, this imposes a serious risk fer an ideal compromise between comprehensive appli- for foot ulceration [56]. cability and simplicity. The reviewers of classification systems usually require each system to include their own particular facet. For example, when the UT sys- 9.8 The classification of ulceration tem was reviewed by Levin [61], he noted that the site of ulceration was missing, despite the fact that this has The most widely used and validated foot ulcer been shown to be an uncertain predictor of outcome. classification systems are the Wagner [57], the A good classification system would seem to require University of Texas (UT) [58, 59] and the PEDIS [60] some allowance for patient factors and the inclusion systems. of a deformity index, particularly in relation to ulcera- The best known system – the Wagner – labels pation in association with Charcot feet (see Section 9.14). tients with risk factors but no ulcer ‘grade 0’, the At present, most of the current classifications force the ‘at-risk’ foot. The classification divides foot ulcers into user to become totally foot-centred at the expense of five categories. the patient as a whole. Whilst this is not likely to create problems in multidisciplinary practice, it is a possible • Grade 1 are superficial ulcers limited to the dermis. cause of fragmented care where the foot clinic is sep- • Grade 2 are transdermal ulcers with exposed tendon arate from diabetology and other support. Addressing or bone, without osteomyelitis or abscess. the social and diabetes factors of patients is likely to improve foot ulcer outcomes [6], and this is particularly • Grade 3 are deep ulcers with osteomyelitis or abscess true of the elderly living alone. formation. • Grade 4 is applied to the feet, with localized gangrene confined to the toes or forefoot. 9.9 The At-risk foot

• Grade 5 applies to feet with extensive gangrene. The mainstay of risk reduction must lie with foot care education and the amelioration of other risk factors, The UT system uses the Wagner grades 1 to 3, but if present (see Table 9.1). Footcare education should = to each adds a stage: A no infection or ischaemia; B be concise and repeated regularly in order to have the = = infection, no ischaemia; C ischaemia no infection; maximum effect on patient behaviour [62]. Video pre- = and D both infection and ischaemia. sentations have been shown to be effective at imparting A significant problem with the Wagner classification knowledge about foot care [63], but should not sup- is that it does not differentiate between those grade 1–3 plant one-to-one or small-group education. The main ulcers which are associated with arterial insufficiency. aspects of foot care education include the need for Such ulcers might be expected to heal less well. Nei- regular, at least once-daily, inspection of the feet for ther does it differentiate those grade 1 and 4 ulcers new lesions, and the need to have shoes measured each which are significantly infected, and which might also time they are acquired. These are two aspects which be expected to have a poorer prognosis. Despite this, appear, from experience, to be regularly overlooked the Wagner classification has been shown to provide an in the majority of patients with ulcers. Although ed- accurate guide to the risk of amputation in a number of ucation is the potential saviour of the diabetic foot, studies, and remains the standard by which other classi- there is a considerable body of evidence to suggest fications have to be judged. The most successful recent that whilst knowledge about diabetic foot problems system is the UT, which uses depth and ischaemia as may increase, attitudes to – and compliance with – the its main classification criteria, and therefore is able to necessary care may remain unchanged [64]. predict the progression from ulceration to amputation The limitations of current education and preventive with some accuracy [58]. In a comparative study, the methods are highlighted by the number of patients Texas system demonstrated some advantages over the who have recurrent ulceration, even in specialist clinics 9.10 SUPERFICIAL ULCERS: WAGNER/UT GRADE 1 121

Table 9.1 At-risk groups for diabetic foot ulceration. Table 9.2 General principles of foot care education. Patients with: 1. Target the level of information to the needs of the • a history of previous ulceration patient. Those not at risk may require only general • peripheral neuropathy advice about foot hygiene and shoes. • peripheral vascular disease 2. Assess the ability of the patient to understand and perform the necessary components of footcare. If this • limited joint mobility is limited, then the spouse or carer should be involved • bony deformities at the beginning of the process. • diabetic nephropathy (especially on dialysis) 3. Suggest a positive approach to foot care with ‘dos’ rather • visual impairment than ‘don’ts’ as the principle of active rather than passive foot care is more likely to be successful and acceptable • a history of alcohol excess to the patient: • patients who live alone • Inspect the feet daily • Report any problems immediately • elderly patients • Have your feet measured every time new shoes are brought • Buy shoes with a square toe box and laces (see Table 9.2). The lack of perceived vulnerability • Inspect the inside of shoes for foreign objects every in neuropathic patients has been highlighted as one day before putting them on reason for this [65, 66]. Until this point is addressed • Attend a fully trained podiatrist regularly effectively, education programmes may be limited in • Cut your nails straight across, and not ‘rounded’ their success. • Keep your feet away from heat (fires, radiators and hot Hospital shoes are the second line of risk reduction water bottles) and check the bathwater before stepping for those patients with deformity which increases foot into it ulcer risk. The attendance of a dedicated orthotist as • Always wear something on your feet to protect them part of the diabetic foot care team can significantly and never walk barefoot improve shoe acceptability and compliance among pa- 4. Repeat the advice at regular intervals and check that it tients, and reduce recurrence rates in those with healed is being followed. ulcers. 5. Disseminate advice to other family members and other As mentioned repeatedly in this text, the elderly pose health care professionals involved in the care of the particular problems when trying to impart effective patient. foot care advice and strategies. Many are unable to perform routine foot care because of poor eyesight and reduced mobility which make it difficult to inspect the foot, and so a spouse or carer should be taught how more commonly seen in neuroischaemic patients, but to provide foot care. There is a particular problem any pressure point can ulcerate in any patient, particu- when the patient lives alone, especially if they are larly callused plantar areas [67]. Superficial ulcers are partially sighted, and this may be insoluble, despite believed to form when pressure leads to a reduction in home support services. skin blood flow, to autolysis, and to a breakdown in the dermal layer which results in the formation of an 9.10 Superficial ulcers: Wagner/UT ulcer (Figures 9.4 and 9.5). Grade 1 Neuropathy allows diabetic patients to continue to stress the skin, by walking or continuing to wear the Superficial plantar ulcers are predominantly neuro- same tight shoes even after the ulcer has formed. Con- pathic in origin, and form at sites of pressure such tinued walking inevitably leads to deterioration in the as metatarsal heads or plantar prominences, includ- foot. The causative factor, which may be unknown to ing the rocker bottom of advanced midfoot Charcot the patient, may be deduced from the site or nature neuroarthropathy (see Section 9.14). Ulceration of the of the ulcer. In particular, any assessment of the pa- dorsum of clawed toes in shoes that are too shallow at tient with a foot ulcer should pay careful attention to the toe-box or the lateral and medial aspects of toes are their shoes. Bedrest causing heel ulcers, and trauma, 122 CH 09 THE DIABETIC FOOT

Figure 9.4 Heel ulcer due to pressure from resting on unprotected heels in a neuropathic patient.

Figure 9.5 Shoe-induced ulceration of clawed toes in a neuropathic diabetic patient. including the heat of hot-water bottles or inappropri- RCW [70]; however, if the RCW is rendered irremov- ate ‘self-chiropody’, are easily recognized causes of able by, for example, wrapping it with Scotchcast, then ulceration [68]. the efficacy of the device in healing ulcers is equiva- The relief of pressure is the principal mode of man- lent to that of the TCC [71]. It must be stressed to the agement of superficial ulcers, regardless of origin. patient that such devices are supplied only for mini- Metatarsal head ulceration can be unloaded in a va- mal walking, such as to the toilet indoors, and not for riety of ways. Bed rest, with adequate heel protection, ‘trips to the shops’. There is often a problem in the pa- is theoretically the most effective, but is difficult to tient group of elderly men or women who live alone, enforce, carries its own risks and, especially if in hospital, is expensive. For this reason a number of am- especially if they abuse alcohol or have no carers. In bulatory methods of off-loading ulcer sites have been such patients the advice to stop excess walking may devised: these include the total contact cast (TCC: not always be followed. ‘gold standard’) and the more often used removable Shoe-induced dorsal and digital ulcers can be easily cast walkers (RCW) or Scotch-cast boot [69]. Research unloaded by the provision of, or by recommencing the has shown that patients frequently fail to wear the wearing of, appropriately fitting wide and extra-depth 9.10 SUPERFICIAL ULCERS: WAGNER/UT GRADE 1 123

Figure 9.6 ‘Scotch-cast’ boot (right foot) and extra-depth shoe (left foot) used in the prevention and treatment of diabetic foot ulceration reduction system, including whether it is being used and whether the patient is able to rest sufficiently at home [72]. shoes with or without insoles and toe spacers or props managed by aggressive debridement and appropriate as required (Figure 9.6). offloading. In neuroischaemic ulceration (UT grade 1C Callus should be debrided at every clinic visit [69, or D), the additional ischaemic risk and potential for 70], as this not only unloads the plantar ulcer but also foot-threatening infections should encourage the use of encourages healing. The formation of excessive callus long-term antibiotics [74]. is a sign that the patient is still walking. Low levels The choice of antibiotics for those with clinically of callus formation are also seen in neuroischaemic infected ulcers is also difficult, but in general most patients, and this too may be debrided with care. opinion seems to support the use of broad-spectrum Whilst bleeding is a sign that viable tissue has been monotherapy; co-amoxiclav is regularly used in the au- reached during the debridement of neuropathic ulcers, thors’ and many other units, as is clindamycin, which the neuroischaemic ulcer should not be traumatized if is also a useful antibiotic for foot infections. These can possible. be started until the results of cultures are known; this The presence or absence of infection is difficult to should be followed by targeted narrow-spectrum an- determine in a diabetic foot ulcer. Necrosis, slough tibiotics prescribed according to the results of cultures. and erythema are not universal and systemic features X-radiography of the foot is recommended to detect are rare; however, the diagnosis remains a clinical one osteomyelitis in the majority of patients with ulcers [73]. Culture from the ulcer surface is likely to pro- which are either deeper than grade 1 on inspection, duce a mixed growth of dubious significance. Culture have a history of penetrative trauma, are associated from ulcer scrapings during debridement, or better, with swelling and redness, are not healing after a deep surgical debridement, may provide a more re- month, or probe to bone. liable guide to the principal organism responsible for It is important to measure the size of the ulcer in the infection; however, this too may be misleading and order to gauge the progress of healing. The minimum broad-spectrum antibiotics are first choice in the treat- measurement should be the diameter of the ulcer in ment of infected ulcers. The provision of ‘prophylactic’ two planes at right-angles. Tracing the perimeter and/or antibiotics has few advocates, and the case for their use photographs are also useful to measure progress. Fail- is not clear from the currently available published evi- ure of the ulcer to heal should prompt an investiga- dence: it is not recommended [73]. Neuropathic ulcers tion as to the effectiveness of any pressure-relieving which are not overtly infected (UT grade 1A) are best modality. 124 CH 09 THE DIABETIC FOOT If all of these aspects are satisfactory, then the Once an ulcer is healed, that patient is left in the question of vascular insufficiency should be addressed. highest category of all for predicting future ulcer risk. Even in apparently purely neuropathic ulcers there Education, footcare, chiropody, footwear and careful may be an underlying element of vascular impairment follow-up are all necessary in an attempt to prevent which, without correction, might significantly impede the recurrence of foot ulceration. the healing of recalcitrant ulceration. The successful treatment of vascular insufficiency can dramatically 9.11 Deep ulcers: Wagner/UT improve healing rates. Dressings alone will not heal an ulcer without ade- Grades 2 and 3 quate pressure relief. Although the dry dressing is still in common use, a moist wound environment encour- Deep ulcers are usually superficial ulcers that have ages granulation tissue formation [75]. Sadly, a good continued to be traumatized by an insensate patient. evidence base for the efficacy of almost any dressing Continued walking on plantar ulcers or the wearing of is lacking [76]. inappropriate shoes advances the cycle of tissue de- The use of wound-healing factors and biosynthetic struction and enlarges the ulcer cavity. If this process skin replacements has not, to date, lived up to initial ex- continues, it may lead to the involvement of underlying pectations. Autologous, blood-derived, wound-healing tendons (Wagner/UT grade 2) and eventually to bone, factor has been used for many years, particularly in causing osteomyelitis (grade 3). Occasionally, pene- the United States, but adequate controlled studies in trating injuries will cause the primary formation of a diabetic patients are uncommon [77]. Platelet-derived deep abscess. growth factor has been shown to increase the healing X-radiography of the feet should be performed rou- rate and total percentage of healed ulcers compared tinely in all patients with deep ulcers. If osteomyelitis is suspected, but is not apparent on initial plain ra- to placebo [78]. The biosynthetic dermal replacement 99m Dermagraft (Smith and Nephew, UK) has shown some diographs, then further investigations with Tc ra- promise in clinical trials of patients with neuropathic dioisotopes and labelled white cell scanning should diabetic foot ulcers [79]. The main market for these be performed. Computed tomography (CT) and mag- products must, however, remain the difficult-to-heal netic resonance imaging scans are also used in centres ulcer, and in most cases appropriate care and pres- with ready access to such facilities. If the ulcer can be sure relief will achieve healing of superficial ulcers of probed to bone, then this is likely to be complicated neuropathic and neuroischaemic origins. by osteomyelitis in all cases, and therefore empirical treatment has been advocated: however, an inability to probe to bone probably excludes osteomyelitis [80]. 9.10.1 The myth of the non-healing ulcer? As with superficial ulcers, pressure relief remains Many reports have tried to categorize ulcers as healing the mainstay of treatment of deep ulceration, but the and non-healing. It is important to be able to identify treatment of sepsis and aggressive surgical debridement those patients in whom treatment is failing and for is increasingly important. If the patient is systemically whom a new approach should be used. If no objective well, and there is no evidence of spreading infection, measure of ulcer healing is used, then there is no then he or she can often be managed as an outpatient. possibility that such patients will be detected, and The use of total contact casts is, however, contraindi- once again the need for measurement and standardized cated in patients with oedema secondary to deep in- descriptions of ulcers cannot be stressed too highly. fection, owing to the risk of swelling within the cast, It is clear that the primary reasons for failure of the leading to cast trauma. Regular debridement down to neuropathic plantar diabetic foot ulcer to heal are inad- the ulcer base is required. Bleeding points demonstrate equate or inappropriate pressure relief, inadequate de- adequate debridement in the neuropathic foot. In the bridement and infection control, failure to recognize or foot with coexisting ischaemia, debridement should be treat vascular insufficiency, or patient non-compliance. less aggressive, and the patient may require admission It is only when all of these factors have been addressed, for investigation and the treatment of osteomyelitis, including angiography and reconstruction where nec- surgical debridement or intravenous antibiotics. essary, or by the use of irremovable off-loading casts Diabetes control should be optimized if possible, that an ulcer can truly be described as non-healing. as there is evidence to suggest that healing is im- Such ulcers will be rare [72]. paired with poor blood glucose control, although this 9.11 DEEP ULCERS: WAGNER/UT GRADES 2 AND 3 125

Figure 9.7 Ray amputation of second toe and associated metatarsal. does not necessarily mean a need for insulin in a healing has been achieved, subsequent graft failure non-insulin-dependent patient. may not lead to loss of the limb as the vascular Podiatric debridement and conservative care will requirements of healed tissue seem to be lower than lead to the healing of over 60% of such ulcers [81]. that of healing ulcers. If surgical debridement is required then it should aim Slough is commonly seen in neuroischaemic ulcers, to remove all the infected tissue in one operation. This and impedes healing by blocking the formation of gran- may necessitate a partial amputation, commonly of a ulation tissue. Chemical debridement with desloughing metatarsal and associated toe (the ‘ray amputation’; agents such as hydrogels or hydrofoams seems to help see Figure 9.7), which should then heal well if the in the early stages of healing if there is adherent slough. blood supply is adequate (Wagner Grade 3, UT 3B) Necrotic eschar is probably best removed mechani- [82]. The removal of all the infected and/or necrotic cally. An alternative is larval debridement, which has tissue should produce an improvement in the patient’s been effective in removing slough and necrotic debris metabolic state and, in neuropathic patients with ade- from wounds in a number of case reports. This method quate blood supply, even extensive tissue loss will heal. is usually used in neuroischaemic ulcers to debride re- Local operations in neuroischaemic patients can lead calcitrant slough. If larvae are used it may take more to larger non-healing wounds, so it is important to en- than one application to clear the ulcer of slough. Re- sure the vascular status of the patient prior to forefoot cent observations suggest that larvae might even be and other surgery. If vascular disease is suspected from helpful in eradicating MRSA from foot ulcers [84]. the absence of pulses (UT 3D), the site of the ulcer, Antibiotic use should be universal in all deep ulcers failure to heal despite adequate therapy or the presence and again should, at least initially, be broad-spectrum of local gangrene, then a vascular opinion should be in nature until a definitive pathogen is isolated. The obtained before any decision to operate. Where indi- average number of potential pathogens isolated from cated, the restoration of an impaired blood flow may a wound swab is more than two organisms [85]. remove the need for amputation, or at least markedly The deeper the tissue that is cultured, or the growth reduce its scope. of an organism from the bloodstream, increases the Previous concerns about the long-term patency of reliability of the pathogenicity of the isolate. Combina- arterial grafts in diabetic patients should be discounted. tion therapy for initial blind treatment has traditionally There is evidence to show that a successful arterial been ampicillin, flucloxacillin and metronidazole intra- bypass operation has the same graft survival rate in venously, or ciprofloxacin and metronidazole. Clinical diabetic patients as in those without diabetes [83]. Once trial evidence has shown that the use of ciprofloxacin 126 CH 09 THE DIABETIC FOOT and clindamycin as combination therapy in oral or intravenous dosing also seems to be effective. Out- patient treatment might be with these antibiotics or clindamycin alone, which is a useful oral antibiotic for the treatment of mild to moderate infections, and in the long-term treatment of osteomyelitis. Dressings should conform to the cavity left by a deep ulcer. Deep ulcers often have tendons at their bases and should not be allowed to become too dry. Once again, the choice of dressing is rarely based upon any clinical trial. Theoretical concepts would point to the use of a moist wound-healing environment with enough absorbency to deal with wound exudation. Foam dressings are the authors’ current choice as a primary or secondary dressing for such ulcers. If the patient has been admitted, then once the initial infection has been controlled the outpatient care, using off-loading casts or similar, can be restarted. Although the clinical progress of these ulcers is slow, eventually complete healing can be achieved. Regular measurement of the ulcer is important to gauge progress above the usual; ‘It looks better = the same = worse’ notes are commonplace in most clinical records. Failure to improve should again prompt a search for the reasons behind the lack of appropriate healing.

Figure 9.8 Localized gangrene at the end of the great 9.12 Localized gangrene: Wagner toe. Grade 4 the technique of percutaneous transluminal angioplasty Localized gangrene is commonly seen at the ends of using an inflatable balloon is increasingly used, even toes (Figure 9.8) and at the apex of the heel. These are in distal obstructions [87]. Positioned at the site of an regions where there are endarteries with little collateral atheromatous narrowing within an artery, the balloon circulation if a feeder branch artery fails. As well as stretches the vessel, thus splitting the plaque and restor- being a sign of global arterial insufficiency in the foot, ing the luminal area. Re-endothelialization must then and therefore of neuroischaemia, digital necrosis can occur over the fissured plaque, and it is usual to use occur as a result of infection in a purely neuropathic intravenous heparin to prevent thrombotic occlusion foot, leading to an infective vasculitis and digital artery during the first 24 h after angioplasty. In general, the closure [86]. success rate for the recanalization of an arterial occlu- Vascular assessment is mandatory for all patients sion by angioplasty is proportionate to the length of with localized gangrene. No clinical arterial insuffi- stenosis or thrombosis. Recanalization can usually be ciency may be found in patients with toe gangrene achieved in over 90% of short stenoses in appropriately alone, but if treatable arterial insufficiency is found skilled hands [88]. However, even a good technical re- then correction will significantly reduce the amount sult with total recanalization of an occluded vessel does of tissue loss. Angioplasty and proximal reconstruc- not always lead to clinical improvement in the limb if tive surgery is as effective in diabetic patients as the distal run-off is poor, and this may require further non-diabetic patients. However, the vascular disease treatment to the distal vessels. Angioplasty has been of diabetic patients is often below the trifurcation of performed at the level of the tibial and peroneal arter- the popliteal artery. ies since 1982, and may be the therapy of first line in Interventional radiology with angioplasty can now elderly diabetic patients with other medical conditions, be used to tackle tibial and peroneal disease. Today, or in whom a vein is not available for distal bypass. 9.12 LOCALIZED GANGRENE: WAGNER GRADE 4 127 Arterial tears and early thrombotic occlusion are the salvage rates from what are highly specialized surgery main adverse events associated with angioplasty, but and angioplasty. fortunately the incidence of these problems is low. There remains a problem with long-term reocclu- 9.12.1 Indications for revascularization sion. Even in technically successful angioplasties with Many surgeons speak of an aggressive limb salvage good run-off, intimal hyperplasia or recurrence of approach in the management of the ischaemic diabetic native disease lead to reocclusion rates which ap- limb [91]. Such an approach is often – and probably proach 50% overall, depending on the duration of the should always be – based on appropriate patient selec- follow-up. However, these rates are similar in diabetic tion. If the patient is expected to have a reasonable life and non-diabetic patients. Re-angioplasty may be pos- expectancy, then it should be remembered that con- sible, and, even if a vessel does re-stenose or occlude, comitant cardiac and cerebrovascular disease is likely there may have been sufficient duration of improved to result in death in over 50% of vascular disease pa- circulation to facilitate healing or to allow a plane of tients within 5 years, regardless of whether they have tissue viability to establish or even close the lesion diabetes [92]. It is the practice of many American (Figure 9.9). Once a lesion is closed, the blood supply surgeons to attend to these vascular beds at the same requirements may be lower than those for an ulcerated time as, or prior to, revascularization of the lower limb, and limb salvage rates are usually higher than limbs, but such a policy requires greater resources than patency rates in most series. are likely to be available in most state-funded health There is evidence that in iliac vessels, or in situations care systems. In general, the patient must be able to un- where re-stenosis is likely, stenting the artery wall dergo what is usually a lengthy operation, particularly can prevent re-occlusion [89]. Iliac angioplasty with for distal bypass, and any pre-existing lung or cardiac or without stenting can also be employed to increase pathology may limit their ability to withstand the oper- arterial inflow to the limb and improve the chances of a ation, or may limit its effectiveness. If the patient has a lower bypass remaining patent. Reconstructive surgery, low functional capacity with a poor potential for reha- particularly with in situ or reversed saphenous vein as bilitation beyond a wheelchair, then such an approach a conduit, can now be performed at the level of the is not tenable. Similarly, if the patient is neither able dorsalis pedis artery to restore pulsatile flow below the nor motivated to walk, or will not stop smoking, then tibial arteries [90]. The availability of such techniques the graft patency will be jeopardized. is often limited to regional centres, and indeed this If surgery to vessels below the knee is required, then may be appropriate in order to achieve the best limb this must be performed with a vein as the conduit.

Figure 9.9 Gangrene has separated to leave a clean ulcer, which subsequently healed well. 128 CH 09 THE DIABETIC FOOT Synthetic grafts, even with vein cuffs, have such low better secondary patency rates. Regardless of the con- patency rates as to render attempts at below-knee duit used, the long-term patency depends on the flow reconstructive surgery using such materials pointless rate through the graft, which in turn is influenced by the [93]. If the leg veins are varicosed, or have been run-off vessels. In most series the 5-year patency aver- harvested for coronary grafting, then an arm vein can ages 70%, although reoperation and redo angioplasty be used, but this adds to the technical aspects and rates are higher in diabetic patients in some series duration of the operation. If a short length of vein [96]. Despite the predilection for vascular disease to can be found, a popliteal artery to foot bypass may be be multi-level and to affect the infra-popliteal vessels almost as successful as a femoral artery to distal bypass in diabetes, there appears to be no significant difference [94]. Similarly, the patient must have suitable anatomy in patency rates between diabetic and non-diabetic pa- with adequate inflow and a patent foot vessel on which tients. This may be due to patient selection, but there to graft. If the nature and extent of infection and also is some evidence that femoral disease and distal necrosis is such that it encroaches upon the potential disease do not always coexist in diabetic patients. In graft site, then again the likelihood is that the graft will addition, owing to a high coexisting mortality, graft fail. This once again highlights the need for control of patency may exceed the life expectancy of the patient infection. [96]. Any centre wishing to offer reconstructive surgery, and particularly distal surgery, must operate a graft 9.12.3 Distal reconstructive operations surveillance programme in order to assess the clinical progress of patients and to audit the results. During the These operations are all outflow procedures performed follow-up period the other vascular trees, coronaries, to vessels below the popliteal artery. As outlined carotids and the other limb may need attention to above, autologous vein is the only suitable conduit reduce the coexisting morbidity and mortality and to for these procedures, which can limit the suitability of improve patient outcome. many patients for surgery. In general, these are oper- The nature of diabetes as a systemic disorder usually ations performed for limb salvage. The flow rate may implies that, in those patients requiring reconstructive mean that in many cases the graft may have failed surgery, there are other associated complications. This by one year; however, the limb be saved if the lesion is particularly true of the elderly patient with diabetes. has closed. In selected centres, the 5-year limb salvage Typically, the intensive care stay is often longer in rates approach 85% despite a graft patency of only diabetic patients, and the perioperative management 68%, and are at least 50% in unselected British centres of diabetes control, cardiac and renal impairment and [26, 92]. Infection should be treated promptly to pre- radiological investigation require a team approach to vent rapidly spreading gangrene and systemic infection the management of surgery in such patients [95]. leading to a severely ill and toxic patient. The antibiotic regimens outlined above under deep ulcers should also be appropriate for these patients. Well-circumscribed, 9.12.2 Proximal arterial reconstruction localized, usually digital, necrosis with viable tissue These operations are divided into inflow procedures, borders can often be left to separate undisturbed; this is usually aorto-iliac surgery, where synthetic graft ma- usually termed ‘auto-amputation’. The wound left be- terials are normally used, and where, because of high hind should then be treated as a neuroischaemic ulcer flow rates, the graft patency is excellent. For aorto- in the usual manner, and will usually heal well. bifemoral grafts the 5-year patency rate is commonly More extensive or spreading necrosis in a toxic pa- over 85%. The patency of aorto-bifemoral grafts is tient, particularly if there is no reversible arterial lesion, the same in both diabetic and non-diabetic patients, may require primary amputation. This decision should although because of associated cardiovascular disease be taken only after review by a vascular surgeon, as the overall patient survival rates are lower in diabetic arterial reconstruction or angioplasty can markedly im- patients, but not usually significantly so [92]. prove the level at which the amputation stump is viable. Reconstructive surgery below the inguinal ligament The remaining foot of an amputee is at an exceed- is usually referred to as an outflow procedure; the usual ingly high risk of ulceration and further surgery. Gen- operation is the femoro-popliteal bypass graft around a eral aftercare should be as for other ulcers, but with superficial femoral occlusion. Synthetic graft materials particular attention to the intact foot. A partial ampu- can be used for these operations, but vein grafts have tation of a toe or ray leads to biomechanical changes 9.14 THE DIABETIC CHARCOT FOOT 129 within the foot which are often very different from and cerebrovascular disease, and treatment for these normal and frequently produces new pressure points at conditions – including aspirin, lipid modification and risk of ulceration. Transmetatarsal or Lisfranc amputa- blood pressure control – should also be addressed dur- tions are often very poorly functioning amputations in ing the follow-up period. diabetic patients. Amputation, at whatever level, results in special orthotic needs that must be addressed by the footcare team. Insoles and orthoses all require careful 9.14 The diabetic Charcot foot and regular review to ensure that they are functioning correctly in order to reduce the significant reulceration The devastating effects of Charcot neuroarthropathy in and amputation rate of diabetic amputees. the diabetic foot, including mortality, have been well described [98–101]. Diabetes is now believed to be the leading cause of Charcot neuroarthropathy in the devel- 9.13 Extensive gangrene: Wagner oped world [102], with 80% of patients who develop Grade 5 Charcot neuroarthropathy having a known duration of diabetes of over 10 years. The long duration of diabetes Extensive necrosis of the foot is due to arterial occlu- prior to the initiation of the Charcot process probably sion and failure of arterial inflow, and usually presents reflects the degree of neuropathy that is usually present with multiple areas of necrosis, generally in the con- in these patients. Autonomic neuropathy appears to be text of the neuroischaemic foot. Primary amputation is a universal finding in diabetic Charcot patients [103]. the usual treatment for extensive gangrene; however, The duration of diabetes appears to be more important the extent of amputation can sometimes be reduced than age alone, but this is compounded in type 2 dia- by pre-amputation arterial reconstructive surgery. For betic patients, who frequently have a long prodromal this reason, the counsel of perfection is that a vascular disease duration prior to diagnosis. assessment should be performed in all patients prior The initiating event of the Charcot process is often to amputation. Either femoropopliteal or similar by- a seemingly trivial injury, which may result in a minor pass operations might improve the viability of a distal periarticular fracture or in a major fracture, despite the stump, or convert an above-knee to a below-knee am- inability of the patient to recall the injury in many putation. Again, this may not always be possible in cases. Following this there is a rapid onset of swelling, diabetic patients, because the arterial disease is often an increase in temperature in the foot, and often an below the popliteal trifurcation, and if the necrosis ex- ache or discomfort. The patient may have noticed a tends beyond the dorsalis pedis artery it will preclude change in the shape of the foot; others have described distal bypass. the sensation, or the sound, of the bones crunching Metabolic and infection control should be attended as they walk. The blood supply to the Charcot foot to as a priority, as these patients are often very ill ow- is always good; indeed, there are case reports of the ing to the toxic effects of the necrotic tissue burden. In Charcot process starting in patients following arterial addition, coexistent coronary and cerebral vascular dis- bypass surgery [104]. It is assumed that autonomic ease often makes the anaesthetic choice difficult, and neuropathy plays a part in the increased vascularity regional anaesthesia is commonly used for amputation of bone, possibly by increased arteriovenous shunting surgery in diabetic patients. Close cooperation between [105], and this increases osteoclastic activity, resulting the medical, surgical and anaesthetic teams is likely to in the destruction, fragmentation and remodelling of produce the best survival outcomes for these patients. bone. It is these processes which, if left untreated, If the patient survives the immediate perioperative lead to the characteristic patterns of deformity in the period, then the mortality rate in patients following Charcot foot, including the collapse of the longitudinal major amputation is >50% at one year. Care of the re- and transverse arches leading to a rocker bottom foot maining foot is particularly important to prevent further (see Figures 9.10 and 9.11). Recent research into the amputation, and this usually results in confinement to pathogenesis of this condition has focused on the a wheelchair. Significant improvements in preservation potential role if inflammation and pro-inflammatory of the remaining limb can be achieved if the patient cytokines [102, 106]. returns to the diabetic foot clinic for follow-up after Charcot neuroarthropathy passes from this acute amputation [97]. The patient is likely to die from other phase of development through a stage of coales- major vessel problems, particularly coronary artery cence, in which the bone fragments are reabsorbed, the 130 CH 09 THE DIABETIC FOOT Intervention must be made in the earliest phase to prevent subsequent deformity and to reduce the risk of amputation [108]. Radiographs of the foot should be performed to make the initial diagnosis (Figure 9.12). The characteristic appearances of bone destruction, fragmentation, loss of joint architecture and new bone formation should be determined. The confirmation of Charcot neuroarthropathy can be made through bone scans, and CT or MRI scans, but this is usually not required in the majority of clinical settings. Management of the Charcot foot has always been difficult, and varies from the expectant to the markedly interventional [109]. The first principles of management are rest and freedom from weight-bearing. Non-weight-bearing is useful to reduce the activity, but weight-bearing frequently restarts when walking is recommenced. In the United States, in particular, the practice of prolonged (one year or more) immobilization in a plaster of Paris cast is the usual treatment. The total-contact cast is usually the method employed, but this requires frequent changes as the oedema reduces. Plaster casting will stabilize the foot but, again, whilst casting reduces activity initially, when the plaster is finally removed after 6–12 months the acute destructive process may Figure 9.10 Anteroposterior view of sole of a Charcot restart. Surgical fusion of the joints of the foot in foot, showing a plantar prominence which has ulcerated. their anatomical positions has usually met with little success during the active phase. oedema lessens, and the foot cools. It then enters the Surgery may still be used, for example to remove a stage of reconstruction, in which the final repair and plantar prominence once the process has finally settled regenerative modelling of bone takes place to leave [78, 110]. The end of the active phase can be assessed a stable, chronic Charcot foot [107]. The time course by following skin temperature and radiographic change of these events is variable but is often up to a year. [111]. In the United Kingdom, total-contact casting is

Figure 9.11 Charcot neuroarthropathy: lateral X-radiograph showing destruction of the talus and mid-foot. 9.15 CONCLUSIONS 131

Figure 9.12 Magnetic resonance image of the talus and mid-foot of the patient in Figure 9.11. Note the bilateral Charcot changes. Such changes are often difficult to interpret, even by experienced radiologists. still the mainstay of treatment in the active phase. The 9.15 Conclusions Scotch-cast boot (see Figure 9.6) can also be used to rest the active Charcot foot, and is particularly useful to The diabetic foot syndrome is a significant cause of provide pressure redistribution of a rocker bottom foot morbidity and mortality in elderly diabetic patients. with an ulcer at its apex. Casting is usually continued However, by recognizing the known risk associations, for several weeks after the temperature differential be- and taking measures to reduce their effect, the inci- ◦ tween the active and contralateral foot is below 1.5 C. dence of foot ulceration can be significantly reduced. The overall role of surgery in the quiescent Charcot If, in turn, foot ulceration is managed in a systematic foot was reviewed by Salamon and Saltzman [112]. and appropriate manner then the incidence of ampu- As yet, there is no definitive treatment aimed tations because of ulceration can be significantly re- at the underlying inflammation or overactivity of duced. This is the ultimate goal in treating diabetic osteoclasts in the active destructive phase of Charcot foot problems. Clear evidence of the success of a mul- neuroarthropathy. Two clinical studies, including a tidisciplinary approach should lead to its adoption more randomized placebo-controlled trial, of the use of widely than is currently the case. intravenous pamidronate have now been performed in acute Charcot neuroarthropathy. In patients with acute destructive-phase Charcot neuroarthropathy, treatment References with intravenous bisphosphonate caused a rapid 1. WHO=IDF (1990) Diabetes care and research in Eu- resolution of symptoms and signs, including foot rope: the St Vincent Declaration. Diabetic Medicine, temperature, and a marked improvement in the 7, 360. biochemical markers of bone turnover, particularly 2. Specialist UK Workgroup Reports (1996) St Vincent alkaline phosphatase concentrations [113, 114]. Such and improving diabetes care: report of the Diabetic therapy should, therefore, be considered in addition to Foot and Amputation Subgroup. Diabetic Medicine, the use of rest and casting outlined above. 13 (Suppl. 4): S27–42. 132 CH 09 THE DIABETIC FOOT

3. SIGN (1997) Management of Diabetic Foot Dis- 17. Thomson FJ, Masson EA and Boulton AJM (1992) ease. Scottish Intercollegiate Network Guideline, Ed- Quantitative vibration perception testing in elderly inburgh. people: an assessment of variability. Age and Aging, 4. McIntosh A, Peters J, Young R et al. (2003) Preve- 21, 171–4. ntion and Management of Foot Problems in Type 2 18. Young MJ, Breddy JL, Veves A and Boulton AJM diabetes: Clinical Guidelines and Evidence. Sheffield, (1994) The use of vibration perception to predict University of Sheffield. (See also www.nice.org.uk/ diabetic neuropathic foot ulceration: a prospective guidance/index ref CG10.) study. Diabetes Care, 17, 557–60. 5. Krishhan S, Nash F, Baker N, Fowler D and Ray- 19. Booth J and Young MJ (2000) Differences in perfor- man G (2008) Reduction in diabetic amputations over mance of commercially available 10 g monofilaments. 11 years in a defined UK population: benefits of Diabetes Care, 23, 984–8. multi-disciplinary team work and continuous prospec- 20. Abbott CA, Carrington AL, Ashe H et al. (2002) The tive audits. Diabetes Care, 31, 99–101. North West Diabetes Foot Care Study: incidence of, 6. Boulton AJM, Vileikyte L, Ragnarson-Tennvall G and and risk factors for, new diabetic foot ulceration in Apelqvist J (2005) The global burden of diabetic foot a community based patient cohort. Diabetic Med, 19, disease. Lancet, 366, 1719–24. 377–84. 7. Thomson FJ and Masson EA (1992) Can elderly 21. Cavanagh PR and Ulbrecht JS (2008) The biomechan- diabetic patients co-operate with routine foot care? ics of the foot in diabetes mellitus. In: JS Bowker and Age and Aging, 21, 333–7. MA Pfeifer (eds) Levin and O’Neal’s The Diabetic 8. Cavanagh PR, Young MJ, Adams JE, Vickers KL Foot. Mosby, Philadelphia, 7th edition, pp. 115–84. and Boulton AJM (1994) Radiographic abnormalities 22. Vinik AI, Maser RE, Mitchell BD and Freeman in the feet of neuropathic diabetic patients. Diabetes R (2003) Diabetic autonomic neuropathy. Diabetes Care, 17, 201–9. Care, 26, 1553–79. 9. Young MJ, Adams JE, Anderson GF, Boulton 23. Purewal TS, Goss DE, Watkins PJ and Edmonds AJM and Cavanagh PR (1993) Medial arterial ME (1995) Lower limb venous pressure in diabetic calcification in the feet of diabetic patients and neuropathy. Diabetes Care, 18, 377–81. matched non-diabetic control subjects. Diabetologia, 24. Flynn MD and Tooke JE (1992) Aetiology of dia- 36, 615–21. betic foot ulceration: a role for the microcirculation? 10. Reiber GE, Pecararo RE and Koepsell TD (1992) Diabetic Medicine, 9, 320–9. Risk factors for amputation in patients with diabetes 25. Sandeman DD, Shore AC and Tooke JE (1992) mellitus. Annals of Internal Medicine, 117, 97–105. Relation of skin capillary pressure in patients with 11. Boulton AJM, Malik RA, Arezzo JC and Sosenko JM insulin-dependent diabetes mellitus to complications (2004) Diabetic somatic neuropathies. Diabetes Care, and metabolic control. New England Journal of 27, 1458–86. Medicine, 327, 760–4. 12. Young MJ, Boulton AJM, Macleod AF, Williams 26. Simms M (2006) Peripheral vascular disease and DRR and Sonksen PH (1993) A multicentre study reconstruction. In: AJM Boulton, PR Cavanagh and of the prevalence of diabetic peripheral neuropathy G Rayman (eds) The Foot in Diabetes. John Wiley in the United Kingdom hospital clinic population. and Sons, Chichester, 4th edn, pp. 250–64. Diabetologia, 36, 150–4 27. Van derFeen C, Neijens FS, Kanters SD et al. 13. Boulton AJM, Vinik AI, Arezzo JC et al. (2005) (2002) Angiographic distribution of lower extremity Diabetic neuropathies: a statement by the American atherosclerosis in patients with and without diabetes. Diabetes Association. Diabetes Care, 28, 956–62. Diabetic Med, 19, 366–70. 14. Young MJ (1998) Capsaicin as topical therapy for 28. Janka HU, Stadl E and Mehnert H (1980) Peripheral painful diabetic neuropathy. The Diabetic Foot, 1, vascular disease in diabetes mellitus and its relation 147–50. to cardiovascular risk factors: screening with Doppler 15. Young MJ and Matthews CF (1998) Screening for ultrasonic technique. Diabetes Care, 3, 207–13. neuropathy – can we achieve our ideals? The Diabetic 29. Jensen T, Borch-Johnsen K, Kofoed-Enevoldsen A Foot, 1, 22–5. and Deckert T (1987) Coronary heart disease in young 16. Boulton AJM, Armstrong DG, Albert SF et al. (2008) Type 1 (insulin-dependent) diabetic patients with and Comprehensive foot examination and risk assessment: without diabetic nephropathy: incidence and risk fac- a report of the task force of the foot care interest tors. Diabetologia, 30, 144–8. group of the American Diabetes Association, with 30. Gibbons GW and Freeman D (1987) Vascular evalua- endorsement by the American Association of Clinical tion and treatment of the diabetic. Clinics in Podiatric Endocrinologists. Diabetes Care, 31, 1679–85. Medicine and Surgery, 4, 377–81. 9.15 CONCLUSIONS 133

31. Orchard TJ and Strandness DE (1993) Assessment of impaired microvascular response to injury in the dia- peripheral vascular disease in diabetes. Diabetes Care, betic foot. Diabetologia, 32, 736–9. 16, 1199–209. 47. MacFarlane IA, Benbow SJ, Chan AW, Bowsher D 32. American Diabetes Association (2003) Peripheral ar- and Williams G (1993) Diabetic peripheral neuropa- terial disease in people with diabetes (Consensus thy: the significance of plantar foot temperatures as Statement). Diabetes Care, 26, 3333–41. demonstrated by liquid crystal contact thermography. 33. Hiatt WR, Regensteiner JG, Hargaten ME et al. Diabetic Medicine, 10 (Suppl. 1), P104. (1990) Benefit of exercise conditioning for patients 48. Masson EA, Hay EM, Stockley I, Veves A, Betts RP with peripheral arterial disease. Circulation, 81, and Boulton AJM (1989) Abnormal foot pressures 602–9. alone may not cause ulceration. Diabetic Medicine, 34. McAllister FF (1976) The fate of patients with inter- 6, 426–8. mittent claudication managed conservatively. Ameri- 49. Van Schie CH, Abbott CA, Vileikyte L, Shaw JE, can Journal of Surgery, 132, 593–5. Hollis S and Boulton AJM. (2000) A comparative 35. Larkin JG and Frier BM (1986) Limited joint mobility study of the podotrack, a simple semi-quantitative and Dupuytrens contracture in diabetic, hypertensive plantar pressure measuring device, and the optical and normal populations. British Medical Journal, 292, pedobarograph in the assessment of pressures under 1494. the diabetic foot. Diabetic Med, 16, 154–9. 36. Goodfield MJB and Millard LG (1988) The skin in 50. Veves A, Van Ross ERE and Boulton AJM diabetes mellitus. Diabetologia, 31, 567–75. (1992) Foot pressure measurements in diabetic and 37. Rosenbloom AL, Silverstain JM, Lezotte DC, non-diabetic amputees. Diabetes Care, 15, 905–7. Richardon K and McCallum M (1981) Limited joint 51. Knowles A, Westwood B, Young MJ and Boulton mobility in childhood diabetes indicates increased AJM (1993) A retrospective study to assess the out- risk for microvascular disease. New England Journal come of diabetic ulcers that have been dressed with of Medicine, 305, 191–4. Granuflex and other dressings. In: Proceedings of the 38. Campbell RR, Hawkins SJ, Maddison PJ and Reckless Joint Meeting of the Wound Healing Society and the JPD (1988) Limited joint mobility in the diabetes European Tissue Repair Society, Amsterdam, August mellitus. Annals of the Rheumatic Disease, 44, 93–7. 1993, P 68. 39. Delbridge L, Perry P, Marr S, Arnold N, Yue DK, 52. Pecoraro RE, Ahroni JH, Boyko EJ and Stensel VL Turtle JR and Reeve TS (1988) Limited joint mo- (1991) Chronology and determinants of tissue re- bility in the diabetic foot: relationship to neuropathic pair in diabetic lower-extremity ulcers. Diabetes, 40, ulceration. Diabetic Medicine, 5, 333–7. 1305–13. 40. Fernando DJS, Masson EA, Veves A and Boulton 53. Louie TJ, Gartlett JG and Tally FP (1976) Aerobic AJM (1991) Relationship of limited joint mobility to and anaerobic bacteria in diabetic foot ulcers. Annals abnormal foot pressures and diabetic foot ulceration. of Internal Medicine, 85, 461–3. Diabetes Care, 14, 8–11. 54. Delbridge L, Appleberg M and Reeves TS (1983) Fac- 41. Young MJ, Cavanagh PR, Thomas G, Johnson MM, tors associated with the development of foot lesions Murray H and Boulton AJM (1992) The effect of in the diabetic. Surgery, 93, 78–82. callus removal on dynamic plantar foot pressures in 55. Young RJ, Zhou YQ, Rodriguez E, Prescott RJ, Ew- diabetic patients. Diabetic Medicine, 9, 55–7. ing DJ and Clark BF (1986) Variable relationship 42. Veves A, Fernando DJS, Walewski P and Boulton between peripheral somatic and autonomic neuropa- AJM (1991) A study of plantar pressures in a diabetic thy in patients with different syndromes of diabetic clinic population. The Foot, 1, 89–92. polyneuropathy. Diabetes, 35, 192–7. 43. Cavanagh PR, Sims DS, Jr and Sanders LJ (1991) 56. Fernando DJS, Hutchison A, Veves A, Gokal R and Body mass is a poor predictor of peak plantar pressure Boulton AJM (1991) Risk factors for non-ischaemic in diabetic men. Diabetes Care, 14, 750–5. foot ulceration in diabetic nephropathy. Diabetic 44. Boulton AJM (2004) The diabetic foot: from art to Medicine, 8, 223–5. science. Diabetologia, 47, 1343–53. 57. Young MJ (2000) Classification of ulcers and its rel- 45. Rayman G, Willams SA, Spencer PD, Smaje LH, evance to management. In: AJM Boulton, PR Ca- Wise PH and Tooke JE (1986) Impaired microvascular vanagh and H Connor (eds) The Foot in Diabetes, response to minor skin trauma in Type 1 diabetes. 3rd edn. John Wiley & Sons, Chichester. British Medicine Journal, 292, 1295–8. 58. Lavery LA, Armstrong DG and Harkless LB (1996) 46. Walmsley D, Wales JK and Wiles PG (1989) Reduced Classification of diabetic foot wounds. Journal of Foot hyperaemia following skin trauma: evidence for an and Ankle Surgery, 35, 528–31. 134 CH 09 THE DIABETIC FOOT

59. Oyibo S, Jude EB, Tarawneh I et al. (2001) The effects 73. Lipsky BA (2008) New developments in diagnosing of ulcer size and site, patient age, sex and type and and treating diabetic foot infections. Diabete Metab duration of diabetes on the outcome of diabetic foot Res Rev, 24 (Suppl. 1): S66–71. ulcer. Diabetic Med, 18, 133–8. 74. Foster A, McColgan M and Edmonds M (1989) 60. Schaper NC (2004) Diabetic foot ulceration classifi- Should oral antibiotics be given to ‘clean’ foot ul- cation system for research purposes: a progress report cers with no cellulites? Diabetic Medicine, 15 (Suppl. on criteria for including patients in research studies. 2), A27 (abstract). Diabete Metab Res Rev, 20 (Suppl. 1), S90–5. 75. Porter M. (1999) Making sense of dressings. Wound 61. Levin ME (1998) Classification of diabetic foot Management, 2, 10–12. wounds. Diabetes Care, 21, 681. 76. Knowles EA (2006) Dressings: is that an 62. Barth R, Campbell LV, Allen S, Jupp JJ and Chisholm evidenced-base? In: AJM Boulton, PR Cavanagh and DJ (1991) Intensive education improves knowledge, G Rayman (eds) The Foot in Diabetes. John Wiley compliance, and foot problems in Type 2 diabetes. & Sons, Chichester, 4th edn, pp. 186–97. Diabetic Medicine, 8, 111–17. 77. Krupski WC, Reilly LM, Perez S, Moss KM, 63. Knowles EA, Kumar S, Veves A, Young MJ, Fer- Crombleholme PA and Rapp JH (1991) A prospec- nando DJS and Boulton AJM (1992) Essential ele- tive randomized trial of autologous platelet-derived ments of footcare education are retained for at least a wound healing factors for the treatment of chronic year. Diabetic Medicine, 9 (Suppl. 2), S6. nonhealing wounds: A preliminary report. Journal of 64. Vileikyte L (2006) Psychological and behavioural is- Vascular Surgery, 14, 526–36. sues in diabetic foot ulceration. In: AJM Boulton, PR 78. Young MJ (1999) Becaplermin and its role in healing Cavanagh and G Rayman (eds) The Foot in Diabetes, neuropathic diabetic foot ulcers. The Diabetic Foot, 2, 105–7. John Wiley & Sons, Chichester, 4th edn, pp. 132–42. 79. Gentzkow GD, Iwasaki SD, Hershon KS et al. (1996) 65. Stuart L and Wiles PJ (1993) Knowledge and beliefs Use of Dermagraft, a cultured human dermis, to treat towards footcare among diabetic patients: a compar- diabetic foot ulcers. Diabetes Care, 19, 350–2. ison of qualitative and quantitative methodologies. 80. Lavery LA, Armstrong DG, Peters EJ and Lipsky BA Diabetic Medicine, 9 (Suppl. 2), S3. (2007) Probe to bone tests for diagnosing diabetic 66. Stuart L and Wiles PJ (1993) The influence of a learn- foot osteomyelitis: reliable or relic? Diabetes Care, ing contract on levels of footcare. Diabetic Medicine, 30, 270–4. 10 (Suppl 3), S3. 81. Pittet D, Wyssa B, Clavel C, Kursteiner K, Vaucher 67. Murray HJ, Young MJ, Hollis S and Boulton AJM J and Lew PD (1999) Outcome of diabetic foot in- (1996) The association between callus formation, high fections treated conservatively; a retrospective cohort pressures and neuropathy in diabetic foot ulceration. study with long term follow up. Archives of Internal Diabetic Medicine, 13, 979–82. Medicine, 159, 851–6. 68. Boulton AJM (1990) Diabetic foot. Neuropathic in 82. McKeown KC (1994) The history of the diabetic foot. origin? Diabetic Medicine, 7, 852–8. In: AJM Boulton, H Connor and PR Cavanagh (eds) 69. Armstrong DG, Lavery LA, Kimbriel HR, Nixon BP The Foot in Diabetes, 2nd edn. John Wiley & Sons, and Boulton AJM (2003) Activity patterns of patients Chichester, pp. 5–14. with diabetic foot ulceration: patients with active 83. Wolfle KB, Bruijnen H, Loeprecht H et al. (2003). ulceration may not adhere to a standard pressure Graft patency in clinical outcome of femoro-distal off-loading regimen. Diabetes Care, 26, 2595–2597. arterial reconstruction in diabetic and non-diabetic 70. Armstrong DG, Lavery LA, Nixon BP and Boulton patients: results of a multicentre comparative analysis. AJM. (2004) It’s not what you put on, but what you Eur J Vasc Endovasc Surg, 25, 229–34. take off: Techniques for debriding and off-loading the 84. Bowling FL, Salgami EV and Boulton AJM (2007) diabetic foot wound. Clin Infect Dis, 39 (Suppl. 2), Larval therapy: a novel treatment in eliminating S92–9. methacillin-resistant Staphylococcus aureus from di- 71. Katz IA, Harlan A, Miranda-Palma B et al. (2005) abetic foot ulcers. Diabetes Care, 30, 370–1. A randomized trial of two irremovable off-loading 85. Hunt JA (1992) Foot infections are rarely due to a devices in the management of plantar, neuropathic single microorganism. Diabetic Medicine, 9, 749–52. diabetic foot ulcers. Diabetes Care, 28, 555–9. 86. Edmonds M, Foster A, Greenhill M, Sinha J, 72. Cavanagh PR, Ulbrecht JS and Caputo GM (1998) Philpott-Howard J and Salisbury J (1992) Acute The non-healing diabetic foot wound: fact or fic- septic vasculitis not diabetic microangiopathy leads tion? Ostomy Wound Management, 44 (3A Suppl.), to digital necrosis in the neuropathic foot. Diabetic 6S–12S. Medicine, 9 (Suppl. 1), 85. 9.15 CONCLUSIONS 135

87. Bolia A (2006) Interventional radiology in the diabetic 100. Sammarco GJ (1991) Diabetic arthropathy. In: GJ foot. In: AJM Boulton, PR Cavanagh and G Rayman Sammarco (ed.) The Foot In Diabetes. Lea & Febiger, (eds) The Foot in Diabetes, 4th edition, John Wiley Philadelphia. & Sons, Chichester, pp. 238–49. 101. Gazis A, Pound N, Macfarlane RM et al. (2004) 88. Mansell PI, Gregson R and Allison SP (1992) An Mortality in patients with diabetic neuropathic audit of lower limb angioplasty in diabetic patients. osteoarthropathy (Charcot foot). Diabetic Med, 21, Diabetic Medicine, 9, 84–90. 1243–6. 89. Palmaz JC, Laborde JC and Rivera FJ (1992) Stenting 102. Jeffcoate WJ (2008) Charcot neuroarthropathy. Dia- of the iliac arteries with the Palmaz stent: experi- bete Metab Res Rev, 24 (Suppl. 1): S58–61. ence from a multicentre trial. Cardiovascular Inter- 103. Marshall A, Young MJ and Boulton AJM (1993) The ventional Radiology, 15, 291–7. neuropathy of patients with Charcot feet: is there a specific deficit? Diabetic Medicine, 10 (Suppl. 1), 90. Estes JM and Pomposelli FB (1996) Lower extremity 101. arterial reconstruction in patients with diabetes melli- 104. Edelman SV, Kosofsky EM, Paul RA and Kozak GP tus. Diabetic Medicine, 13 (Suppl. 1), S43–57. (1987) Neuro-osteoarthropathy (Charcot’s joints) in 91. Gibbons GW (1994) Vascular surgery: its role in diabetes mellitus following revascularisation surgery: foot salvage. In: AJM Boulton, H Connor and PR three case reports and a review of the literature. Cavanagh (eds) The Foot in Diabetes. John Wiley & Archives of Internal Medicine, 147, 1504–8. Sons, Chichester, pp. 177–90. 105. Edmonds ME, Clarke MB, Newton S, Barrett J and 92. Sigurdsson HH, Macaulay EM, McHardy KC Watkins PJ (1985) Increased uptake of bone ra- and Cooper GG (1999) Long-term outcome of diopharmaceutical in diabetic neuropathy. Quarterly infra-inguinal bypass for limb salvage: are we giving Journal of Medicine, 57, 843–55. diabetic patients a fair deal? Practical Diabetes 106. Jeffcoate WJ, Game F and Cavanagh PR (2005). The International, 16, 204–6. role of pro-inflammatory cytokines in the cause of 93. Cheshire NJW, Wolfe JHN, Noone MA, Davies BA neuropathic osteoarthropathy (acute Charcot foot) in and Drummond M (1992) The economics of femoro- diabetes. Lancet, 366, 2058–61. crural reconstruction for critical leg ischaemia with 107. Eichenholtz SN (1966) Charcot Joints.CharlesC. and without autologous vein. Journal of Vascular Thomas, Springfield, IL. Surgery, 15, 167–75. 108. Gazis A, Macfarlane RM and Jeffcoate WJ (2000) 94. Pomposelli JB, Jepson SJ, Gibbons GW, Campbell Delay in diagnosis of the Charcot foot. Diabetic DR, Freeman DV, Miller A and LoGerfo FW (1991) Medicine, 17 (Suppl. 1), 80. A flexible approach to infra-popliteal vein grafts in 109. Lesko P and Maurer RC (1989) Talonavicular disloca- patients with diabetes mellitus. Archives of Surgery, tions and midfoot arthropathy in neuropathic diabetic 126, 724–9. feet: natural course and principles of treatment. Clin- 95. Hirsch IB and White PF (1988) Medical management ical Orthopaedics, 240, 226. of surgical patients with diabetes. In: ME Levin and 110. Young MJ (1999) The management of neurogenic LW O’Neal (eds) The Diabetic Foot. CV Mosby, St arthropathy – A tale of two Charcots. Diabetes Metabolism Research and Reviews, 15, 59–64. Louis, pp. 423–32. 111. Sanders LJ and Frykberg RG (1991) Diabetic 96. Bartlett FF, Gibbons GW and Wheelcock FC (1986) neuropathic osteoarthropathy: the Charcot foot. Aortic reconstruction for occlusive disease: compa- In: RG Frykberg (ed.) The High Risk Foot in rable results in diabetics. Archives of Surgery, 121, Diabetes. Churchill Livingstone, New York, 1150–3. pp. 297–338. 97. Abbott CA, Carrington AL and Boulton AJM (1996) 112. Salamon M and Saltzman CL (2006) The operative Reduced bilateral amputation rate in diabetic patients: treatment of Charcot neuroarthropathy of the foot and effect of a foot care clinic. Diabetic Medicine, 13 ankle. In: AJM Boulton (ed.) The Foot in Diabetes. (Suppl. 7), S45. John Wiley & Sons, Chichester, pp. 274–84. 98. Sinha S, Munichoodappa CS and Kozak GP. Neu- 113. Selby PL, Young MJ and Boulton AJM (1994) roarthropathy (Charcot joints) in diabetes mellitus. Pamidronate in the treatment of diabetic Charcot neu- Medicine (Baltimore), 51, 191–210. roarthropathy. Diabetic Medicine, 11, 28–31. 99. Cofield RH, Morrison MJ and Beabout JW (1983) 114. Jude EB, Selby PL, Burgess J et al. (2001). Bisphos- Diabetic neuroarthropathy in the foot: patient char- phonates in the treatment of Charcot neuroarthropa- acteristics and patterns of radiographic change. Foot thy: a double-blind randomised controlled trial. and Ankle, 4, 15–22. Diabetologia, 44, 2032–7. 10 Diabetic Neuropathy

Solomon Tesfaye University of Sheffield, Royal Hallamshire Hospital, Sheffield Teaching Hospitals NHS Foundation Trust, Sheffield, UK

involved the examination of 3250 type 1 diabetes pa- Key messages tients, from 16 European countries, found a prevalence rate of 28% for DPN at baseline [2]. The study also • Diabetic neuropathy may affect up to 30% of all showed that, over a 7-year period, about one-quarter people with diabetes, and is the main initiating of type 1 diabetic patients developed DPN, with age, factor for foot ulceration and amputation. duration of diabetes and poor glycaemic control be- • The detection of neuropathy in a patient with ing major factors [3]. The development of neuropathy diabetes should not prevent the usual screening was also associated with potentially modifiable cardio- for other causes of neuropathy. vascular risk factors such as hypertension, hyperlipidaemia, obesity and cigarette smoking [3]. Based on recent epidemiological studies, correlates of DPN in- 10.1 Introduction clude increasing age, increasing duration of diabetes, poor glycaemic control, retinopathy, albuminuria and Diabetic neuropathy is a common complication of dia- vascular risk factors [3]. betes and a cause of considerable morbidity and increased mortality. Diabetic neuropathy encompasses several neuropathic syndromes, the most common of 10.3 Classification of diabetic which is diabetic peripheral neuropathy (DPN), the polyneuropathy main initiating factor for foot ulceration. Some patients with peripheral neuropathy may experience trouble- Figure 10.1 shows a modified clinical classification some neuropathic pain that is difficult to treat. DPN of diabetic polyneuropathy originally suggested by is also associated with autonomic neuropathy that can Thomas [4]. Attempts at classification stimulate involve almost all the systems of the body and may thought as to the aetiology of the various syndromes, have devastating consequences, such as sudden death. and also assist in the planning of management strategy for the patient. 10.2 Epidemiology 10.4 Diabetic peripheral Both, clinic- and population-based studies show sur- neuropathy (DPN) prisingly similar prevalence rates for DPN, as affecting approximately 30% of all diabetic people [1]. The This is by far the most common neuropathic syndrome EURODIAB Prospective Complications Study, which encountered in clinical practice. The sensory symptoms

Focal and Multifocal Polyneuropathy neuropathies

mono- multiple lesions acute autonomic neuropathy ‘mononeuritis sensory multiplex’

amyotrophy radiculopathy

Diabetic peripheral neuropathy Entrapment (DPN) eg median ulnar peroneal (most common syndrome)

Figure 10.1 Classification of diabetic polyneuropathy [4]. start in the toes and then extend to involve the feet It is a great paradox that subjects with painful and legs in a stocking distribution. Once the disease neuropathy may at the same time have a severe absence is established in the lower limbs, the upper limbs also of sensation, the so-called ‘painful-painless foot’ [8]. become involved, with a similar progression that starts DPN is indeed the main initiating risk factor for foot in the fingers. There is an associated autonomic neu- ulceration and amputation. This underpins the need for ropathy that is often detectible by autonomic function careful examination and screening of the feet of all tests; however, clinically overt autonomic neuropathy diabetic people, in order to identify those at risk of de- is less common. veloping foot ulceration. This is particularly important The main clinical presentation of DPN is a sen- in the elderly, as many live alone (are socially isolated), sory loss which the patient may not be aware of; it are immobile and have poor vision. The insensate foot may be described as ‘numbness’ or ‘dead feeling’. is at risk of developing mechanical and thermal in- As the disease advances, motor manifestations such juries, and patients must therefore be warned about as wasting of the small muscles of the hands and limb these and given appropriate advice with regard to foot weakness become apparent. Some patients may expe- care. In those with advanced neuropathy, there may be rience a progressive build-up of unpleasant sensory sensory ataxia. The unfortunate sufferer is affected by symptoms [5] in the lower limbs, including tingling unsteadiness on walking and even falls, particularly in (paraesthesiae), burning pain; shooting pains, lancinat- the elderly and those with poor vision. ing pains, contact pain (allodynia) and aching pain. DPN is usually easily detected by a simple, bed-side Occasionally, the pain can extend above the feet and peripheral neurological examination [9]. For this, the may involve the whole of the legs, and when this is shoes and socks should be removed and the feet exam- the case there is usually involvement also of the upper ined at least annually, but more often if neuropathy is limbs. present. The most common presenting abnormality is Painful diabetic neuropathy is characteristically a reduction or absence of vibration sense in the toes. more severe at night, and often disturbs sleep [6]. It As the disease progresses there is sensory loss in a also has a major impact on functionality (e.g., ability ‘stocking’ and sometimes in a ‘glove’ distribution in- to maintain work), mood and the quality of life. volving all modalities. When there is severe sensory Depressive symptoms as a result of the unremitting loss, proprioception may also be impaired, leading to pain are common [7]. a positive Romberg’s sign. Ankle tendon reflexes are 10.4 DIABETIC PERIPHERAL NEUROPATHY (DPN) 139 lost, and with a more advanced neuropathy the knee Table 10.2 Differential diagnosis of DPN. reflexes are often reduced or absent. Many of these Metabolic sensory modalities, including ankle reflexes, are also • Diabetes decreased with advancing age; consequently, a good discriminator for the presence of DPN in old age is • Amyloidosis the loss of any ‘pin prick’ sensation. • Uraemia Muscle strength is usually normal early during the • Myxoedema course of the DPN. However, with progressive disease • Porphyria there may be generalized muscular wasting, especially • in the small muscles of the hands and feet. Affected Vitamin deficiency (thiamine, B12,B6, pyridoxine) patients may have problems with the fine movements Drugs and chemicals of fingers and in handling small objects. The typical clawing of the toes seen in advanced DPN is due to • Alcohol unopposed (because of wasting of the small muscles of • Cytotoxic drugs (e.g. vincristine) the foot) pulling of the long extensor and flexor ten- • Chlorambucil dons. This results in elevated plantar pressure points • Nitrofurantoin at the metatarsal heads that are prone to callus forma- • tion and foot ulceration. Deformities such as a bunion Isoniazid can form the focus of ulceration and with more ex- Neoplastic disorders treme deformities, such as those associated with Char- cot arthropathy [10], the risk is further increased. As • Bronchial or gastric carcinoma inappropriate footwear is the most common form of • Lymphoma trauma to the neuropathic foot, any clinical assessment should also include an examination of the shoes for Infective or inflammatory poor fit, abnormal wear, and internal pressure areas or • Leprosy foreign bodies. A summary of the clinical assessment • Guillain–Barre´ syndrome of the diabetic patient for DPN is shown in Table 10.1. • Lyme borreliosis Autonomic neuropathy affecting the feet can cause • a reduction in sweating and consequent dry skin that Chronic inflammatory demyelinating polyneuropathy is likely to crack easily, thus predisposing the patient • Polyarteritis nodosa to the risk of infection. The neuropathic foot without Genetic peripheral vascular disease is also warm, due to the presence of arteriovenous shunting and distended • Charcot–Marie–Tooth disease veins [11]. • Hereditary sensory neuropathies

Table 10.1 Clinical assessment for DPN. 10.4.1 Differential diagnosis of DPN History Signs Before attributing the neuropathy to diabetes, other • Sensory symptoms Inspection common causes of neuropathy must first be excluded. • Motor symptoms Reflexes (ankle reflex The absence of other complications of diabetes, rapid unreliable in the elderly) weight loss, excessive alcohol intake and other atypical • Assessment of disability Sensory features in either the history or clinical examination • • Exclude other causes of vibration should direct the physician to search for other causes neuropathy • light touch of neuropathy (Table 10.2). • Pinprick (good discriminator in the elderly) 10.4.2 Acute painful neuropathies • 10 g Monofilament Acute painful neuropathies are transient neuropathic Assess footwear syndromes characterized by an acute onset of pain 140 CH 10 DIABETIC NEUROPATHY (within weeks) in the lower limbs, and are relatively prognosis is good, with usually complete resolution uncommon. Marked neuropathic pain involving both of symptoms within 12 months. The management of limbs is invariably present, and often distressing to the painful symptoms is as used to treat chronic DPN. patient. There are two distinct syndromes, the first of which occurs within the context of poor glycaemic control, and the second with rapid improvement in 10.5 Asymmetrical neuropathies glycaemic control. Focal or asymmetrical neuropathies have a relatively Acute painful neuropathy of poor glycaemic rapid onset, and complete recovery is usual. This con- control trasts with chronic DPN, where there is usually no improvement in symptoms for several years after onset. This occurs usually in type 1 or type 2 diabetic subjects Unlike chronic distal symmetrical neuropathy, asym- with poor glycaemic control, and there is often an as- metrical neuropathies are often unrelated to the pres- sociated severe weight loss [12]. Ellenberg coined the ence of other diabetic complications. Asymmetrical description of this condition as ‘neuropathic cachexia’ neuropathies predominantly affect middle-aged/older [13]. Patients typically experience a persistent burning patients and are more common in men [15]. A careful pain associated with allodynia (contact pain). Although history is therefore essential in order to identify any the pain is most marked in the feet, it often affects the associated symptoms that might point to another cause whole of the lower extremities. As in chronic DPN, the for the neuropathy. pain is typically worse at night, although unremitting pain during day time is also common. The acute-onset distressing pain often results in depression. 10.6 Diabetic amyotrophy In acute painful neuropathies the sensory loss is usually surprisingly mild, or even absent. There are This syndrome was first described by Garland [16], usually no motor signs, although ankle jerks may be ab- who noted the presence of progressive, asymmetri- sent. Nerve conduction studies are also usually normal cal proximal leg weakness and atrophy. Consequently, or mildly abnormal. The temperature discrimination Garland coined the term ‘diabetic amyotrophy’, al- threshold (small fibre function) is however, affected though this condition has also been named ‘proximal more commonly than the vibration perception thresh- motor neuropathy’, ‘femoral neuropathy’or‘plexopa- old (large fibre function). There is complete resolution thy’. Both, type 1 and type 2 diabetic patients aged of symptoms within 12 months, and weight gain is over 50 years are affected. usual with continued improvement in glycaemic con- The patient presents with severe pain which is felt trol with the use of insulin. deep in the thigh, but can sometimes be of burning quality and extend below the knee. The pain is usually Acute painful neuropathy of rapid glycaemic continuous and often causes insomnia and depression control (insulin neuritis) [17]. There is an associated weight loss which can The term ‘insulin neuritis’ is a misnomer, as the sometimes be very severe, and can raise the possi- condition can follow rapid improvement in glycaemic bility of an occult malignancy. On examination, there control with oral hypoglycaemic agents. The author is profound wasting of the quadriceps, with marked has therefore recommended that the term ‘acute weakness in these muscle groups, although the hip flex- painful neuropathy of rapid glycaemic control’ be used ors and hip abductors can also be affected. The thigh to describe this condition [14]. The natural history adductors, glutei and hamstring muscles may also be of acute painful neuropathies is an almost guaranteed involved, and the knee jerk is usually reduced or ab- improvement [14], in contrast to chronic DPN. The sent. Such profound weakness can lead to difficulty in patient presents with burning pain, paraesthesiae, rising from a low chair or climbing stairs. Sensory loss allodynia, often with a nocturnal exacerbation of symp- is unusual, but if it is present it indicates a coexistent toms; depression may also be a feature. There is no DPN. associated weight loss, unlike acute painful neuropathy Other causes of quadriceps wasting, such as nerve of poor glycaemic control. Sensory loss is often mild root and cauda equina lesions and occult malignancy or absent, and there are no motor signs. There is little causing proximal myopathy syndromes (e.g. polymy- or no abnormality on nerve conduction studies. The ocytis), should be excluded. Today, MR imaging of the 10.9 PRESSURE PALSIES 141 lumbosacral spine is mandatory in order to exclude fo- or the abdomen [21]. The pain is usually asymmetrical, cal nerve root entrapment and other pathologies. An and can cause local bulging of the muscle [22]. There erythrocyte sedimentation rate (ESR), X-radiography may also be patchy sensory loss, and other causes of of the lumbar/sacral spine and chest X-radiography and nerve root compression should be excluded. Recovery ultrasound of the abdomen may also be required. Elec- is usually the rule within several months, although trophysiological studies may demonstrate an increased symptoms can sometimes persist for a few years. femoral nerve latency and active denervation of the Some patients presenting with abdominal pain have affected muscles. Levels of cerebrospinal fluid (CSF) undergone unnecessary investigations such as barium protein are often elevated. enema, colonoscopy and even laparotomy, when the The natural history of diabetic amyotrophy is that of diagnosis could easily have been made via a careful gradual recovery, although there is paucity of prospec- clinical history and examination. tive studies. Coppack and Watkins [17] have reported that pain usually starts to settle after about 3 months, and usually settles by 1 year, while the knee jerk is 10.9 Pressure palsies restored in 50% of the patients after 2 years. Recur- rence, on the other hand, is a rare event. Management is 10.9.1 Carpal tunnel syndrome largely symptomatic and supportive; patients should be encouraged and reassured that this condition is likely The patient typically has pain and paraesthesia in the to resolve. Controversy persists as to whether the use hands, but this sometimes radiates to the forearm and is of insulin therapy influences the natural history of this particularly marked at night. In severe cases a clinical syndrome. Some patients benefit from physiotherapy examination may reveal a reduction in sensation in the that involves extension exercises aimed at strengthen- median territory in the hands, and wasting of the mus- ing the quadriceps. The management of pain in prox- cle bulk in the thenar eminence. The clinical diagnosis imal motor neuropathy is similar to that of chronic is easily confirmed by median nerve conduction stud- painful DPN. ies, while treatment involves surgical decompression at the carpel tunnel in the wrist. There is generally a good response to surgery, although painful symptoms 10.7 Cranial mononeuropathies may relapse more commonly than in the non-diabetic population. Third cranial nerve palsy is the most common cranial mononeuropathy. Typically, the patient will present with pain in the orbit, or occasionally with a frontal 10.9.2 Ulnar nerve and other isolated headache [18]. Frequently, ptosis and ophthalmoplegia nerve entrapments is also evident, although the pupil is usually spared The ulnar nerve is also vulnerable to pressure damage [19]. Recovery occurs usually over 3–6 months. The at the elbow, and this results in a wasting of the dor- clinical onset and time scale for recovery, and the sal interossei, particularly the first dorsal interossius. focal nature of the lesions on the third cranial nerve identified at post-mortem examination, suggested an This is easily confirmed by ulnar electrophysiological ischaemic aetiology [20]. It is important to exclude any studies. other cause of third cranial nerve palsy (aneurysm or Rarely, the patient may present with wrist drop tumour) by the use of CT or MR scanning, where the due to radial nerve palsy after prolonged sitting (with diagnosis is in doubt. Fourth, sixth and seventh cranial pressure over the radial nerve in the back of the arms), nerve palsies have also been described in diabetic while unconscious during hypoglycaemia, or asleep subjects. after excessive alcohol intake. In the lower limbs the common peroneal (lateral popliteal) is the most commonly affected nerve, and 10.8 Thoracoabdominal neuropathy this results in foot drop. Unfortunately, complete recovery is not usual. The lateral coetaneous nerve of Diabetic thoracoabdominal neuropathy (also known as the thigh is occasionally also affected with entrapment truncal radiculopathy) is characterized by an acute neuropathy in diabetes. Phrenic nerve involvement in onset pain in a dermatomal distribution over the thorax association with diabetes has also been described. 142 CH 10 DIABETIC NEUROPATHY 10.10 Pathogenesis of diabetic Table 10.3 Proposed hypotheses of peripheral neuropathy diabetic peripheral nerve damage. • Chronic hyperglycaemia Historically, there have been two distinct views with • Nerve microvascular dysfunction regards to the pathogenesis of DPN. The first view • Increased free radical formation involves the metabolic factors [23] that are primarily • important in the pathogenesis of DPN, while the second Polyol pathway hyperactivity view contends that vascular factors [24] determine • Protein kinase C hyperactivity the etiological factors for neuropathy (Table 10.3). • Non-enzymatic glycation Today, however, most authorities agree that the truth • Abnormalities of nerve growth most likely lies between the two cases, and that both metabolic and vascular factors are important. Table 10.4 Clinical consequences of auto- 10.11 Autonomic neuropathy nomic neuropathy. Cardiac autonomic neuropathy Autonomic neuropathy can affect many systems and • sudden death result in significant morbidity (Table 10.4) and mortal- • ity. Autonomic involvement usually has a gradual onset silent ischaemia and is slowly progressive. In the EURODIAB study, • exercise intolerance the prevalence of autonomic neuropathy (defined as the • orthostatic hypotension presence of two abnormal cardiovascular autonomic • foot vein distension/arteriovenous shunting function tests) was 23%, and the prevalence increased with age, duration of diabetes, glycaemic control and Gastrointestinal autonomic neuropathy presence of cardiovascular risk factors, in particular • hypertension [25]. gastroparesis • diarrhoea or constipation 10.11.1 Cardiovascular autonomic neuropathy Bladder hypomotility • Cardiovascular autonomic neuropathy is a serious com- urinary incontinence/retention plication of longstanding diabetes, and causes postural Erectile dysfunction hypotension, changes in peripheral blood flow, and may also be a cause of sudden death. Gustatory sweating

Postural hypotension The management of subjects with postural hypoten- It is now generally accepted that a fall in systolic blood sion poses major problems, and for some patients there pressure of >20 mmHg is considered abnormal [25]. may not be any satisfactory treatment. Current treat- Coincidental treatment with tricyclic antidepressants ments include: (i) the removal of any drugs that might for neuropathic pain, and also with diuretics, may result in orthostatic hypotension, such as diuretics, exacerbate postural hypotension, the chief symptom beta-blockers, anti-anginal agents; (ii) advising patients of which is dizziness on standing. The symptoms to get up from the sitting or lying position very slowly, of postural hypotension can be disabling for some and to cross the legs while doing so; (iii) increasing the patients who may not be able to walk for more than sodium intake to 10 g (185 mmol) per day and fluid a few minutes. Severely affected patients are prone to intake to 2–2.5 l per day (note: care must be exer- unsteadiness and falls. The degree of dizziness does cised in elderly patients with heart failure); (iv) the not appear to correlate with the postural drop in blood use of custom-fitted elastic stockings which extend to pressure. There is increased mortality in subjects with the waist; (v) treatment with fludrocortisone (starting postural hypotension, although the reasons for this are at 100 μg per day) while carefully monitoring urea not fully clear. and electrolytes; and (vi) in severe cases the alpha-1 10.11 AUTONOMIC NEUROPATHY 143 adrenoreceptor agonist, midodrine (or occasionally oc- sometimes the presence of succusion splash, while bar- treotide) may be effective. ium swallow and follow through, and also gastroscopy, may reveal a large food residue in the stomach. Gas- Changes in peripheral blood flow tric motility and emptying studies can sometimes be performed in specialized units, and may help with di- Autonomic neuropathy can cause arteriovenous shunt- agnosis. ing, with prominent veins in the neuropathic leg [11]. Management of diabetic gastroparesis include: opti- The leg vein oxygen tension and capillary pressure are mization of glycaemic control; the use of anti-emetics increased in the neuropathic leg due to sympathetic (metoclopramide and domperidone) and the use of a denervation. Thus, in the absence of peripheral vas- cholinergic agent that stimulates oesophageal motil- cular disease the neuropathic foot is warm, and this ity (erythromycin, which may enhance the activity of may be one of the factors that cause osteopenia associ- the gut peptide, motilin). Gastric electrical stimulation ated with the development of Charcot neuroarthropathy (GES) has recently been introduced as a treatment op- [10]. tion in patients with drug refractory gastroparesis to increase the quality of life by alleviating nausea and Cardiovascular autonomic function tests vomiting frequencies [28]. This service is offered at Five cardiovascular autonomic function tests are now specialist units. widely used for the assessment of autonomic function. Severe gastroparesis causing recurrent vomiting, is These tests are non-invasive, and do not require sophis- associated with dehydration, swings in blood sugar ticated equipment. All that is required is an electrocar- levels and weight loss, and is therefore an indication for diogram machine, an aneroid pressure gauge attached hospital admission. The patient should be adequately to a mouthpiece, a hand-grip dynamometer and sphyg- hydrated with intravenous fluids and the blood sugar momanometer. Reference data for the cardiovascular level should be stabilized by intravenous insulin, while autonomic function tests are listed in Table 10.5 [26]. anti-emetics could be given intravenously. If the course of the gastroparesis is prolonged, then total parenteral 10.11.2 Gastrointestinal autonomic nutrition or feeding through a gastrostomy tube may neuropathy be required. Gastroparesis Autonomic diarrhoea Autonomic neuropathy can reduce oesophageal motility (dysphagia and heart-burn), and cause gastroparesis The patient may present with diarrhoea that tends to be (reduced gastric emptying, vomiting, swings in blood worse at night, or alternatively some may present with sugar levels) [27]. constipation. Both, the diarrhoea and constipation, re- The diagnosis of gastroparesis is often made on spond to conventional treatment. Diarrhoea associated clinical grounds by the evaluation of symptoms, and with bacterial overgrowth may respond to treatment

Table 10.5 Reference values for cardiovascular function tests. Parameter Normal Borderline Abnormal Heart rate tests Heart rate response to standing up (30: 15 ratio) ≥1.04 1.01–1.03 ≤1.00 Heart rate (beats/min) response to deep breathing (maximum ≥15 11–14 ≤10 minus minimum heart rate) Heart rate response to Valsalva manoeuvre (Valsalva ratio) ≥1.21 – ≤1.20 Blood pressure tests Blood pressure response to standing up (fall in systolic BP) ≤10 mmHg 11–29 mmHg ≥30 mmHg Blood pressure response to sustained hand-grip (increase in ≥16 mmHg 11–15 mmHg ≤10 mmHg diastolic BP) 144 CH 10 DIABETIC NEUROPATHY with a broad-spectrum antibiotic such as erythromycin, and an empathetic approach is essential. In general, tetracycline or ampicillin. patients should be allowed to express their symptoms freely without too many interruptions. The psycholog- 10.11.3 Abnormalities of bladder function ical support of the patient’s painful neuropathy is an important aspect of the overall management of the pain. Bladder dysfunction is a rare complication of autonomic neuropathy and may result in hesitancy 10.12.1 Glycaemic control of micturition, an increased frequency of micturition and, in serious cases, with urinary retention associated Today, there is little doubt that good blood sugar con- with overflow incontinence. Such a patient is prone to trol prevents/delays the onset of diabetic neuropathy urinary tract infections. Ultrasound scan of the urinary [30]. In addition, painful neuropathic symptoms may tract, in addition to urodynamic studies, may be also improved by improving metabolic control, if nec- required. Treatment manoeuvres include mechanical essary with the use of insulin in type 2 diabetes, al- methods of bladder emptying by applying suprapubic though evidence from controlled trials is missing [31]. pressure, or the use of intermittent self-catheterization. Thus, the first step in the management of painful neu- Anti-cholinesterase drugs such as neostigmine or ropathy is an attempt to improve glycaemic control, peridostigmine may be useful. Long-term indwelling where appropriate. Clearly, in elderly patients where catheterization may be required in some cases, but there is a significant risk of hypoglycaemia, this may this unfortunately predisposes the patient to urinary not be advisable. tract infections and long-term antibiotic prophylaxis may be required. 10.12.2 Tricyclic compounds Tricyclic compounds are regarded as one of the 10.11.4 Gustatory sweating first-line treatment agents for painful diabetic neuropathy [29], with a number of double-blind clinical Increased sweating usually affecting the face, and of- trials having confirmed their effectiveness beyond ten brought about by eating (gustatory sweating), can any doubt. As these drugs have unwanted adverse be very embarrassing to patients. Unfortunately there side effects such as drowsiness, anticholinergic side is no totally satisfactory treatment for gustatory sweat- effects such as dry mouth, and dizziness due to ing, although the anticholinergic drug poldine may be postural hypotension in those patients with autonomic useful in a minority of patients. neuropathy, all patients should be commenced on imipramine or amitriptyline at a low dose level 10.12 Management of painful (10–25 mg, taken before bed), with the dose gradually diabetic neuropathy being titrated as necessary up to 100 mg per day. Caution should be taken in elderly patients and in The treatment scenario for painful neuropathy is less those with cardiovascular disease; in general, these than satisfactory, as currently available treatment ap- compounds are also best avoided in patients with documented cardiovascular disease, as an increased proaches may not completely abolish the pain [29]. A mortality has been reported. The mechanism of action careful history and examination of the patient is essen- of tricyclic compounds in improving neuropathic pain tial in order to exclude other possible causes of leg is not fully understood; however, their effect does pain, such as peripheral vascular disease, prolapsed not seem to be mediated via their antidepressant intervertebral discs, spinal canal stenosis and corda properties, as they appear to be effective even in those aquina lesions [29]. Unilateral leg pain should arouse patients in depressed mood [32]. a suspicion that the pain may be due to lumbar-sacral nerve root compression. These patients may well need 10.12.3 Serotonin noradrenaline reuptake to be investigated with a lumbar-sacral magnetic res- inhibitors onance imaging (MRI). Other causes of peripheral neuropathy include an excessive alcohol intake and vi- Serotonin noradrenaline reuptake inhibitors (SNRIs), tamin B12 deficiency. Where pain is the predominant such as duloxetine, relieve pain by increasing the symptom, the quality and severity should be assessed; synaptic availability of 5-hydroxytryptamine (5-HT) neuropathic pain may be disabling in some patients, and noradrenaline (norepinephrine) in the descending 10.12 MANAGEMENT OF PAINFUL DIABETIC NEUROPATHY 145 pathways that are inhibitory to pain impulses. Duloxe- pain [38]. Another opioid, oxycodone slow release, has tine is now also a first line-agent, and is licensed for the also been shown to be effective in the management treatment of painful diabetic neuropathy. The efficacy of neuropathic pain [39]. Recently, combinations of of duloxetine in painful neuropathy has been investi- morphine and gabapentin [40], and oxycodone and gated in three identical trials [33], from which pooled gabapentin [41], were found to be more effective than data have shown daily doses of both 60 mg and 120 mg either drug alone on its own in the management of to be effective in relieving painful symptoms. Treat- diabetic neuropathic pain. ment was found to be effective within a week and this was maintained for 12 weeks. Duloxetine is contraindicated in patients with liver disease. Although nausea is 10.12.9 Management of disabling painful the most common side effect, it is usually self limiting. neuropathy not responding to pharmacological treatment 10.12.4 Anticonvulsants Neuropathic pain can sometimes be extremely se- Older anticonvulsants, including sodium valproate and vere, and interfere significantly with the patient’s carbamazepine, are effective but tend to have more sleep and daily activities. Unfortunately, those patients adverse side effects. Gabapentin (and more recently who are not helped by conventional pharmacologi- pregabalin) administered at 300–600 mg per day [34] cal treatment may pose a major challenge, as they have also been found effective, and can serve as first-line agents for the management of painful diabetic may become severely distressed and are occasionally neuropathy. Adverse side effects include dizziness, somnolence and peripheral oedema.

Painful diabetic neuropathy 10.12.5 Intravenous lignocaine Intravenous lignocaine at a dose of 5 mg kg−1 body- Depending on contra-indications weight, with a further 30 min with a cardiac monitor and co-morbidities in situ, has also been found to be effective in relieving neuropathic pain for up to 2 weeks [35]. This form of Alpha-2-delta SNRI treatment is useful in subjects that have severe pain agonist (pregabalinor TCA (duloxetine) which will not respond to the above-described agents; gabapentin) however, the patient must be brought into the hospital for a few hours during treatment. If pain is inadequately controlled and depending on contra-indications 10.12.6 Alpha-lipoic acid SNRI or TCA or alpha- Infusion of the antioxidant alpha-lipoic acid, at a dose TCA or SNRI alpha-2-delta 2-delta of 600 mg per day, both orally and intravenously, has agonist agonist also been found to be useful in reducing neuropathic pain [36]. If pain is still inadequately controlled 10.12.7 Lacosamide Lacosamide represents another promising anticonvul- sant for the treatment of painful DPN. Whilst, in a Opioid agonist as monotherapy, followed by combination therapy if pain control is phase II study, lacosamide was found to be beneﬁcial still inadequate in relieving painful DPN, phase III studies are now required to assess the situation [37]. Key: TCA = tricyclic antidepressant; SNRI = selective serotonin noradrenaline re-uptake inhibitor 10.12.8 Opiates The opiate derivative tramadol (50–100 mg, four times Figure 10.2 Proposed treatment algorithm for painful daily) has been found effective in relieving neuropathic diabetic neuropathy. (Adapted from Ref. [43]). 146 CH 10 DIABETIC NEUROPATHY wheelchair-bound. Such patients may respond to elec- 10. Rajbhandari SM, Jenkins R, Davies C and Tesfaye trical spinal cord stimulation which relieves both back- S. Charcot neuroarthropathy in diabetes mellitus. Di- ground and peak neuropathic pain [42]. This form of abetologia 2002; 45; 1085–96. treatment is particularly advantageous, as the patient 11. Ward JD, Simms JM, Knight G, Boulton AJM and San- does not have to take any other pain-relieving medica- dler DA. Venous distension in the diabetic neuropathic tions, with all their adverse side effects. foot (physical sign of arterio-venous shunting). J Roy Soc Med 1983; 76: 1011–14. 12. Archer AG, Watkins PJ, Thomas PJ, Sharma AK and 10.12.10 Treatment algorithm Payan J. The natural history of acute painful neuropathy in diabetes mellitus. J Neurol Neurosurg Psychiatr A recent consensus meeting carefully evaluated the 1983; 46: 491–6. trial evidence for the various pharmacological treat- 13. Ellenberg M. Diabetic neuropathic cachexia. Diabetes ments for painful DPN, and suggested a treatment 1974; 23: 418–23. algorithm [43] (Figure 10.2). It must be emphasized 14. Tesfaye S, Malik R, Harris N, Jakubowski J, Mody C, that, as all of the pharmacological agents used have Rennie IG and Ward JD. Arteriovenous shunting and adverse side effects, the drugs must be commenced at proliferating new vessels in acute painful neuropathy of the smallest doses and gradually increased in order to rapid glycaemic control (insulin neuritis). Diabetologia minimize such problems. This is particularly important 1996; 39: 329–35. 15. Matikainen E and Juntunen J. Diabetic neuropathy: in elderly patients who are prone to falls and often live Epidemiological, pathogenetic, and clinical aspects alone. with special emphasis on type 2 diabetes mellitus. Acta Endocrinol Suppl. (Copenh) 1984; 262: 89–94. 16. Garland H. Diabetic amyotrophy. Br Med J 1955; 2: References 1287–90. 17. Coppack SW and Watkins PJ. The natural history of 1. Ziegler D. Diagnosis, staging and epidemiology of femoral neuropathy. Q J Med 1991; 79: 307–13. diabetic peripheral neuropathy. Diab Nutr Metab 1994; 18. Asbury AK, Aldredge H, Hershberg R and Fisher CM. 7: 342–8. Oculomotor palsy in diabetes mellitus: a clinicopatho- 2. Tesfaye S, Stephens L, Stephenson J, Fuller J, Platter logical study. Brain 1970; 93: 555–7. ME, Ionescu-Tirgoviste C and Ward JD. The preva- 19. Leslie RDG and Ellis C. Clinical course following lence of diabetic neuropathy and its relation to gly- diabetic ocular palsy. Postgrad Med J 1978; 54: 791–2. caemic control and potential risk factors: the EURO- 20. Dreyfuss PM, Hakim S and Adams RD. Diabetic oph- DIAB IDDM Complications Study. Diabetologia 1996; thalmoplegia. Archives of Neurology and Psychiatry 39: 1377–84. 1957; 77: 337–49. 3. Tesfaye S, Chaturvedi N, Eaton SEM, Witte D, Ward 21. Ellenberg M. Diabetic truncal mononeuropathy - a new JD and Fuller J. Vascular risk factors and diabetic clinical syndrome. Diabetes Care 1978; 1: 10–13. neuropathy. New Engl J Med 2005; 352: 341–50. 22. Boulton AJM, Angus E, Ayyar DR and Weiss R. Dia- 4. Thomas PK. Metabolic neuropathy. J Roy Coll Phys betic thoracic polyradiculopathy presenting as abdom- (Lond) 1973; 7: 154–74. inal swelling. Br Med J 1984; 289: 798–9. 5. Tesfaye S. Diabetic neuropathy: achieving best prac- 23. Stevens MJ, Feldman EL, Thomas T and Greene tice. Br J Vasc Dis 2003; 3: 112–17. DA. (1998) Pathogenesis of diabetic neuropathy. In: 6. Tesfaye S and Price D. (1997) Therapeutic approaches A. Veves (ed.) Contemporary Endocrinology: Clinical in diabetic neuropathy and neuropathic pain. In: AJM Management of Diabetic Neuropathy. Humana Press, Boulton (ed). Diabetic Neuropathy. Marius Press, Carn- Totowa NJ, pp. 13–48. forth, pp. 159–81. 24. Cameron NE, Eaton SE, Cotter MA and Tesfaye S. Vas- 7. Quattrini C and Tesfaye S. Understanding the impact of cular factors and metabolic interactions in the patho- painful diabetic neuropathy. Diabetes Metab Res Rev genesis of diabetic neuropathy. Diabetologia 2001; 44: 2003; 19 (Suppl. 1), S1–8. 1973–88. 8. Ward JD. The diabetic leg. Diabetologia 1982; 22: 25. Witte DR, Tesfaye S, Chaturvedi N, Eaton SEM, Kem- 141–7. pler P, Fuller JH and the EURODIAB Prospective 9. Eaton SEM and Tesfaye S. Clinical manifestations and Complications Study Group. Risk factors for cardiac measurement of somatic neuropathy. Diabetes Reviews autonomic neuropathy in Type 1 diabetes mellitus. Di- 1999; 7: 312–25. abetologia 2005; 48: 164–71. 10.12 MANAGEMENT OF PAINFUL DIABETIC NEUROPATHY 147

26. Ewing DJ, Martyn CN, Young RJ and Clarke BF. The 36. Zeigler D, Hanefeld M, Ruhnau KJ, et al. Treatment value of cardiovascular autonomic function tests: ten of symptomatic diabetic peripheral neuropathy with years experience in diabetes. Diabetes Care 1985; 8: anti-oxidant alpha-lipoic acid: a 3-week multicentre 491–8. randomised controlled trial (ALADIN Study). Dia- 27. Horowitz M and Fraser R. Disordered gastric motor betologia 1995; 38: 1425–33. function in diabetes mellitus. Diabetologia 1994; 37: 37. Rauck RL, Shaibani A, Biton V, Simpson J and Koch 543–51. B. Lacosamide in painful diabetic peripheral neuropa- 28. Lin Z, Forster J, Sarosiek I and McCallum RW. Treat- thy: a phase 2 double-blind placebo-controlled study. ment of diabetic gastroparesis by high-frequency gas- Clin J Pain 2007; 23: 150–8. tric electrical stimulation. Diabetes Care 2004; 27 (5): 38. Harati Y, Gooch C, Swenson M, Edelman S, Greene 1071–6. D, Raskin P, Donofrio P, Cornblath D, Sachdeo R, 29. Tesfaye S and Kempler P. Painful diabetic neuropathy. Siu CO and Kamin M. Double-blind randomized trial Diabetologia 2005; 48: 805–7. of tramadol for the treatment of the pain of diabetic 30. Diabetes Control and Complications Trial Research neuropathy. Neurology 1998; 50 (6): 1842–6. Group. (1995) The effect of intensive diabetes therapy 39. Watson CP, Moulin D, Watt-Watson J, Gordon A on the development and progression of neuropathy. and Eisenhoffer J. Controlled-release oxycodone re- Ann Int Med 122: 561–8. lieves neuropathic pain: a randomized controlled trial 31. Boulton AJM, Drury J, Clarke B and Ward JD. Contin- in painful diabetic neuropathy. Pain 2003; 105 (1-2): uous subcutaneous insulin infusion in the management 71–8. of painful diabetic neuropathy. Diabetes Care 1982; 5: 40. Gilron I, Bailey JM, Tu D, Holden RR, Weaver DF and 386–90. Houlden. Morphine, gabapentin, or their combination 32. Max MB, Culnane M, Schafer SC et al. Amitriptyline for neuropathic pain. N Engl J Med. 2005; 352 (13): relieves diabetic neuropathy pain in patients with nor- 1324–34. mal or depressed mood. Neurology 1987; 37: 589–96. 41. Hanna M, O’Brien C and Wilson MC. Prolonged- 33. Goldstein DJ, Lu Y, Detke MJ, Lee TC and Iyengar S. release oxycodone enhances the effects of existing Duloxetine vs. placebo in patients with painful diabetic gabapentin therapy in painful diabetic neuropathy pa- neuropathy. Pain 2005; 116 (1-2): 109–18. tients. Eur J Pain 2008; 12 (6): 804–13. 34. Freeman R, Durso-Decruz E and Emir B. Efﬁcacy, 42. Tesfaye S, Watt J, Benbow SJ, Pang KA, Miles J and safety and tolerability of pregabalin treatment of MacFarlane IA. Electrical spinal cord stimulation for painful diabetic peripheral neuropathy: ﬁndings from 7 painful diabetic peripheral neuropathy. Lancet 1996; randomized, controlled trials across a range of doses. 348: 1698–701. Diabetes Care 2008 31 (7): 1448–54. 43. Jensen TS, Backonja MM, Hernandez Jimenez S, Tes- 35. Kastrup J, Angelo H, Petersen P, Dejgard˚ A and Hilsted faye S, Valensi P and Ziegler D. New perspectives J. (1986) Treatment of chronic painful neuropathy with on the management of diabetic peripheral neuropathic intravenous lidocaine infusion. Br Med J 292: 173. pain. Diab Vasc Dis Res 2006; 3: 108–19. 11 Erectile Dysfunction

Tamas´ Varkonyi´ 1 and Peter Kempler2 1I. Department of Medicine, University of Szeged, Szeged, Hungary 2I. Department of Medicine, Semmelweis University, Budapest, Hungary

used to define ED, and also to the lack of systematic Key messages stratification by age. In the Massachusetts Male Aging Study, ED to some degree was found in 52% of • Several complications and concomitant condi- adult men between the ages of 40 and 70 years [4]. tions of diabetes are regarded as risk factors of Approximately 35–75% of men with diabetes mellitus erectile dysfunction. have ED [5], and in these cases the ED has been shown • The cardiovascular risk of diabetic men with to occur 5–10 years earlier than in age-matched control erectile dysfunction must be assessed by the subjects [6]. In a study of 541 men with diabetes application of the Princeton guidelines. conducted at a community-based clinic, the prevalence • Trials with phosphodiesterase type 5 inhibitors in diabetic patients have demonstrated an improve- of ED was seen to increase progressively with age, ment of erectile dysfunction, without reporting from 6% in men aged 20–24 years to 52% in those any increased mortality or high frequency of aged 55–59 years [7]. cardiovascular adverse events. Among a cohort of patients with longstanding type 1 diabetes mellitus (≥10 years), ED was reported in 1.1% of men aged 21–30 years, in 55% of men 11.1 Introduction aged 50–60 years, and in 75% of those aged >60 years [8]. The prevalence and risk factors of diabetic 11.1.1 Definition of erectile dysfunction complications were assessed in 3250 type 1 diabetic subjects in 31 centers as part of the EURODIAB IDDM In 1993, the idiom of impotence was replaced by the Complications Study; the data relating to ED are listed term erectile dysfunction (ED) which, according to the in Table 11.1 [9]. The range of responses given to NIH Consensus Development Panel, is defined as the various questions regarding ED varied enormously ‘...persistent inability to achieve or maintain an erec- among centres, from 2% to 85%. The similarity of tion sufficient for satisfactory sexual intercourse’ [1]. patterns of answers at some centres suggested that cultural and personal attitudes did not allow patients 11.1.2 Prevalence of erectile dysfunction and their doctors to communicate in such a way that The reported prevalence of ED among the general the true prevalence could be ascertained. This might population ranges from 19 to 52% [2, 3], this span indicate that, in some centres, it was not possible for being due most likely to differences in the criteria patients with ED to discuss their problems [9].

Table 11.1 The prevalence of erectile dysfunction in sion, diabetes mellitus, hypercholesterolaemia, hyper- the EURODIAB IDDM complications study. triglyceridaemia and obesity) were examined in 570 community-dwelling men, aged 30–69 years (mean age Complication Proportion of patients (range) (%) 46 years) [11]. After an average follow-up period of 25 years, incident ED was evaluated using the five-item Problems with intercourse 16 (2–35) International Index of Erectile Function. The results Problems obtaining an erection 16 (2–85) Problems sustaining an erection 18 (2–83) showed that age, body mass index (BMI) and hyperc- No erections at night or in the 35 (23–84) holesterolaemia were each significantly associated with morning an increased risk of ED. Overall, one in five men had three or more risk factors and were at a 2.2-fold increased risk of ED [11]. 11.2 Erectile dysfunction: An When assessing the risk factors of ED in men observable marker of diabetes visiting outpatient clinics, 86% of those with ED had one or more chronic disease. ED was present in 70–3% mellitus? of men with coronary heart disease, in 67.8% of those with hypertension, in 78% of those with diabetes, and Using a nationally representative managed care claims in 70.5% of patients with psychiatric diseases [12]. database from 51 health plans and 28 million mem- Some aspects of the metabolic syndrome are asso- bers in the United States, a retrospective cohort study ciated with ED, although it is not clear whether the was conducted to compare the prevalence rates of risk of ED when all components of the syndrome are diabetes mellitus between men with ED (285 436) present is greater than the risk associated with each and men without ED (1 584 230) between 1995 and of its components independently [5]. The results of 2001 [10]. Logistic regression models were used to cross-sectional studies have shown that men with a isolate the effect of ED on the likelihood of hav- BMI >28.7 have a 30% higher risk for ED than those ing diabetes mellitus with adjustment for age, region with a normal BMI (≥25) [13]. The prevalence of and seven concurrent diseases. The diabetes melli- obesity and associated vascular risk factors in men re- tus prevalence rates were 20.0% in men with ED, porting symptoms of ED is remarkably high [14, 15]. and 7.5% in men without ED; with adjustment for In order to study whether men with ED were more age, region and concurrent diseases, the odds ratio of likely to have hypertension than those without ED, having diabetes mellitus between men with ED and hypertension prevalence rates were assessed among without ED was 1.60 (p < 0.0001). With adjustment 285 436 men with ED, and also in 1 584 230 men for regions and concurrent diseases, the age-specific without ED [16]. The prevalence of hypertension was odds ratios ranged from 2.94 (p < 0.0001, age 26–35 seen to be 41.2% in men with ED, and 19.2% in = years) to 1.05 (p 0.1717, age 76–85 years). In sum- those without ED. After controlling for subject age, mary, men with ED were more than twice as likely census region and nine concurrent diseases, the odds to have diabetes mellitus as men without ED. The ratio (OR) was 1.383 (p < 0.0001), which implied that authors concluded that ED should be considered as the odds for men with ED to have hypertension were an observable marker of diabetes mellitus, strongly 38.3% higher than the odds for men without ED. ≤ so for men aged 45 years, and likely for men aged Longitudinal data from the Massachusetts Male Ag- ≥ 46–65 years, but was not a marker for men aged 66 ing Study (MMAS) showed that smoking at baseline years. doubled the likelihood of moderate or complete ED occurring 8 years later (24% for smokers versus 14% in non-smokers) [17]. 11.3 Risk factors of erectile Overall, epidemiologic studies have provided pow- dysfunction erful support for the role of cardiovascular risk factors in ED [17, 18], with diabetes, hypertension and treated As part of a prospective epidemiological survey – the heart disease each having been demonstrated as ma- Rancho Bernardo Study – common coronary artery jor independent risk factors for ED and cardiovascular disease (CAD) risk factors (age, smoking, hyperten- disease [17, 18]. 11.5 ERECTILE DYSFUNCTION: A FIRST SIGN OF CARDIOVASCULAR DISEASE? 151 11.4 Pathophysiology of ED in • Peripheral vascular disease resulting in reduced ar- diabetes mellitus terial and arteriolar inflow [21]. • Advanced glycation end products, leading to in- Among all chronic diseases, diabetes mellitus appears creased amounts of reactive oxidizing substances and to have the most substantiated and characterized asso- reduced NO production [22, 23]. ciation with ED (Figure 11.1). • Failed neural transmission from the spinal cord due Data obtained from a recent study suggested that ED to diabetic neuropathy and a reduced production of occurs at a young age in diabetic men compared to neuronal NO synthase-reduced levels of neuronal non-diabetic individuals. Nicolosi et al. found that the NO release to the cavernosal smooth muscle [24]. frequency of ED in diabetic men (aged 45–49 years) was similar to that in non-diabetic men (aged 70+ • Hypogonadotrophic hypogonadism [25]. years). ED in patients with diabetes has a multifactorial aetiology [19], the main contributory factors being summarized as: 11.5 Erectile dysfunction: A first sign of cardiovascular disease? • Hyperglycaemia and increasing age, leading to glycation of the elastic fibres and failure of relaxation Evidence is accumulating in favour of a link between of the corpora cavernosa [20]. ED and coronary artery disease. As noted above, the • Multiple drug treatments associated with ED, includ- common risk factors for atherosclerosis are prevalent ing diuretics and beta-blockers. in patients with ED, and the extent of ED has been related to the number and severity of the risk factors • Dyslipidaemia. themselves [26]. The prevalence of ED is increased • Endothelial dysfunction of the sinusoidal endothelial in patients with vascular diseases, such as CAD [27], cells, resulting in decreased nitric oxide (NO) release diabetes [28], cardiovascular disease [29], hypertension and impaired vasodilatation. [16] and peripheral artery disease [30]. Consequently,

Causal factors and consequencies of ED in diabetes

Diabetes-associated causes Consequences

Hyperglycaemia Depression

Concomitant Impaired quality of hypertension life

Accelerated large-vessel Risk of poor atherosclerosis Erectile dysfunction cardiovascular condition

Microvascular disease Need to modify previous therapy

Dyslipidaemia Introduction of first-or second-line therapy Autonomic neuropathy

Figure 11.1 Causal factors and consequences of ED in diabetes. 152 CH 11 ERECTILE DYSFUNCTION a new onset or a progressive decline in erectile func- 11.7 Other causes of erectile tion represent a ‘red flag’ for possible cardiovascular problems, with ED serving as a ‘barometer’ of over- dysfunction all vascular fitness. Questions concerning ED should Although the vascular diseases and other factors listed be included in the review of systems for men aged above should be considered as the most important over 30 years, with ED being used as a ‘hook’ to gain aetiological factors of ED, certain other causes should compliance regarding the aggressive management of perhaps also be mentioned. existing cardiovascular disease. A proportion of men will care more about their current ED than about a future risk of myocardial infarction or stroke. ED is not a 11.7.1 Erectile dysfunction and depression clinically silent condition; rather, it may become man- Quite often, depression will accompany aging, and in- ifest before other cardiovascular conditions do, and yet deed an association has been identified between dia- appear ‘silent’ to the health care provider, with many betic neuropathy and depressive symptoms among di- men remaining silent about the symptom unless they abetic patients [32]. The treatment of ED in depressed are asked. patients has been shown to improve the depressive symptoms. 11.6 The artery size hypothesis: A Neurogenic causes of ED include: macrovascular link between • Radical pelvic surgery ED and CAD • Pelvic/spinal cord injury • Currently, evidence is accumulating that ED should be Multiple sclerosis or demyelinating conditions considered a vascular disorder. Common risk factors • Neuropathies for atherosclerosis are frequently found in association • with ED, and the latter is frequently reported in vas- Pudendal nerve injury cular syndromes. Finally, a similar early impairment • Stroke, Alzheimer’s and Parkinson’s disease of endothelium-dependent vasodilatation and late obstructive vascular changes has been reported in both The role of afferent and efferent neuropathies is shown ED and other vascular diseases. Recently, a patho- in Table 11.2. physiological mechanism has been proposed to explain the link between ED and coronary disease, termed the ‘artery size hypothesis’ [31]. Given the systemic nature Table 11.2 Erectile dysfunction: the role of efferent and of atherosclerosis, all major vascular beds should be af- afferent neuropathies. fected to the same extent; however, symptoms rarely become evident at the same time. This difference in Efferent neuropathy Afferent (somatosensory) the rate of occurrence of different symptoms has been neuropathy suggested to result from the different sizes of the arter- • A result of disruption or • Also called ‘dorsal ies which supply the different vascular beds that allow dysfunction of parasympa- nerve impotence’ a larger vessel to better tolerate the same extent of thetic efferent neural path- • Caused by polysensory plaque compared to a smaller vessel. According to this ways neuropathies secondary • hypothesis, because the penile arteries are smaller in di- Causes include spinal cord to diabetes, vitamin B12 ameter than the coronary arteries, patients with ED will injury, multiple sclerosis, deficiency and lead poi- seldom have concomitant symptoms of CAD, whereas radical prostatectomy, pe- soning those with coronary disease will frequently complain of ripheral neuropathy, other ED. Available clinical evidence appears to support this neurological diseases hypothesis [31]. 11.9 TREATMENT OF ERECTILE DYSFUNCTION IN DIABETIC PATIENTS 153

11.7.2 Erectile dysfunction and drug use of diabetes. According to the ﬁndings of a detailed review of 428 charts of male diabetic patients from Erectile dysfunction is very frequently associated with 10 general practices in the UK, the prevalence of ED drug use; typical drugs responsible for ED include: was 53%. However, ED was documented only in 8% antihypertensives [thiazide diuretics, beta-blockers, of cases, while the patient had been informed about calcium-channel blockers, angiotensin-converting possible treatment options in only 1% of cases. enzyme (ACE) inhibitors], antidepressants [tricyclic The most important diagnostic tests to be performed antidepressants, selective serotonin reuptake in- in men with ED are summarized in Table 11.4. A hibitors (SSRIs)], anti-arrhythmics (e.g. digoxin), comprehensive assessment of the patient’s sexual and anti-androgens, H2 receptor antagonists (e.g. cime- medical history is of utmost importance; the key com- tidine), recreational/abuse agents, cigarette smoking ponents are listed in Table 11.5. and cocaine or marijuana. 11.9 Treatment of erectile 11.8 Differentiation between dysfunction in diabetic patients organic and psychogenic erectile dysfunction The unmet need for the treatment of ED in Europe is very high, according to the observations of The Erectile Psychogenic ED is common is diabetic patients, the Dysfunction Observational Study. In fact, 66% of pa- aetiology of the condition including anxiety, depres- tients in nine European countries had experienced ED sion, sexual phobias, stress, direct inhibition of spinal symptoms for one year or longer when seeking treat- centers and excessive sympathetic outﬂow. The key ment [33]. Diabetic patients seeking treatment for ED differences between psychogenic and organic ED are have a greater severity of disease, a less-impaired sex- detailed in Table 11.3. ual desire, and a more organic than functional origin, which suggests that the need for medical care among diabetic patients with ED could be even greater than 11.8.1 Diagnostic testing in their non-diabetic counterparts [34]. Hence, those The prognosis of ED is very important, as ED is doctors participating in diabetes care have an impor- considered as one of the most neglected complications tant role to motivate their patients in this behaviour:

Table 11.3 The classiﬁcation of erectile dysfunction: psychogenic or organic?

Aetiology Psychogenic Organic • Onset • Sudden onset • Gradual onset • Initiation • Complete immediate loss • Incremental progression • Morning erection • Morning erections present • Lack of morning erections • Presence or absence under • Varies with partner and cir- • Lack of erections under most sexu- certain circumstances cumstance ally stimulating circumstances

Table 11.4 Erectile dysfunction: diagnostic testing.

Mandatory or routine tests Recommended tests Specialized tests • Comprehensive history (sexual, • Testosterone (total, free or • Nocturnal penile tumescence medical, drug, and psychosocial) bioavailable) • Vascular studies • Focused physical examination • Fasting glucose and serum lipids 154 CH 11 ERECTILE DYSFUNCTION

Table 11.5 Diagnosis of erectile dysfunction: a com- social phobias play an important role in the pathogene- prehensive history. sis of ED [40]. It is important to note that this treatment option is not efficient in all diabetic men because there Sexual history Medical history are several severe organic aetiological factors involved • • Erectile insufficiency Rule out comorbid condi- in the pathogenesis of ED. • Altered patient or part- tions As hypogonadism is a frequent cause of ED in di- ner sexual desire • Atherosclerotic risk fac- abetes and obesity, the replacement of androgens is • Ejaculation tors and vascular disease a possible first-line option in these patients. Besides • monotherapy in many cases, this treatment is recom- • Orgasm Use of medications/recreational drugs or smoking mended as a combination for those patients who failed • Partner sexual function • History of surgeries or phosphodiesterase type 5 (PDE5) therapy alone [41]. • Sexually induced geni- pelvic/perineal trauma The route of testosterone administration can be differ- tal pain ent [42]; for example, transdermal application involves • Depressive symptoms the use of a gel or scrotal and non-scrotal patches. The intramuscular injection of testosterone cyprionate or enanthate ensures a long-acting effect on ED. Im- in a recent survey, just over half of all men (50.3%) plantable pellets are also available, but these have not with diabetes considered that their doctor should rou- yet been used in a high number of patients. Orally ad- tinely ask about their sexual health [19]. The therapy of ministered testosterone is quickly metabolized by the ED in diabetic patients should be approached as first-, liver, which prevents the achievement of sufficiently second- and third-line options. high blood levels. The alkylated androgen preparations that are currently available in the United States are 11.9.1 First-Line therapy generally not recommended because of poor androgen effects, adverse lipid changes and hepatic side effects. A crucial part of the first-line therapy is the modifi- In Europe, testosterone undecanoate has also a lim- cation of lifestyle and concomitant drug therapy. The ited use. Concerns regarding an increased incidence of lifestyle should be modified by smoking cessation, lim- prostatic cancer, together with cardiovascular risks of itation or avoidance of alcohol, the application of car- testosterone replacement therapy, have led to a cau- bohydrate and fat restrictions in the diet, the achieve- tious consideration of indication and a strict follow-up ment of weight loss, an initiation of regular physical of the patients because no long-term safety data have exercise and an insurance of an adequate sleep [35, been available until very recently [43]. 36]. As diabetic patients take a high number of drugs, the consideration of any potential associations between current treatments and ED should be applied among the PDE5 Inhibitors first-line interventions. Among antihypertensive drugs, One of the most important first-line therapeutic ap- beta-blockers, thiazide diuretics, spironolactones and proaches for ED is an inhibition of the PDE5 enzyme. clonidine should be avoided [37] or replaced (if possi- This results in an amplification of the natural release ble) by ACE inhibitors or angiotensin receptor block- of intracavernosal endothelial and neuronal nitric ox- ers, as these drugs have no harmful effect on erectile ide in response to sexual stimulation. An abnormally function. Some of centrally acting drugs, such as SS- low level of cGMP is increased by delaying the degra- RIs or tricyclic antidepressants, are frequently adminis- dation of this molecule; this in turn leads to decreased tered to diabetic patients – particularly to treat chronic intracellular Ca levels, producing smooth muscle re- painful neuropathy. The detrimental influence of this laxation in the corpus cavernosum and an increased therapy on ED must be taken into account [38]. Drugs blood flow [44]. As ED is closely associated with car- which interfere with the endocrine system may alter diac risk factors in diabetic patients, PDE5 inhibitors sexual function; consequently, histamine H2 receptor should be very carefully considered among those pa- antagonists and metoclopramide must be carefully rec- tients with cardiac risk. Notably, the risk categories of ommended for diabetic men [39]. patients must be taken into account before these drugs Psychosocial counselling and education should be are administered. On this basis, two international con- considered as parts of any first-line treatment, as psy- sensus conferences on sexual activity and cardiac risk chological problems such as anxiety, depression and were convened at Princeton University in 1999 and 11.9 TREATMENT OF ERECTILE DYSFUNCTION IN DIABETIC PATIENTS 155

Table 11.6 The risk categories of the Princeton guidelines. Low-risk patient Indeterminate-risk patient High-risk patient Asymptomatic, <3 cardiac risk ≥3 major cardiac risk factors Unstable or refractory angina factors Controlled hypertension Moderate, stable angina Uncontrolled hypertension Mild, stable angina Recent myocardial infarction Recent myocardial infarction (>2 weeks, <6 weeks) (<2 weeks) Post-successful state of coronary Left ventricular dysfunction Left ventricular dysfunction revascularization (NYHA class II) (NYHA class III-IV) Uncomplicated past myocardial Non-cardiac sequelae of High-risk arrhythmias infarction atherosclerotic diseases Mild valvular disease Hypertrophic obstructive and other cardiomyopathies Left ventricular dysfunction Moderate/severe valvular (NYHA class I) disease

2004, mainly in order to categorize these patients [45, stratification. As a first step of this process, an as- 46]. Three risk categories of patient were identified, sessment of sexual function and initial cardiovascular based on several very important reports of the consen- evaluation is recommended. The second step is to ini- sus panels: tiate or resume sexual activity or to indicate treatment for ED in those patients at low risk. At this phase, the • Patients in the low-risk group included: asymp- patients at high risk should be stabilized by cardio- tomatic patients with fewer than three cardiac risk logic treatment before sexual activity is considered, or factors; patients in whom the blood pressure is well the treatment of sexual dysfunction is recommended. controlled with more than one antihypertensive med- Patient follow-up and reassessment at regular intervals ication; patients with mild, stable angina or past are recommended in all cases. The third step involves revascularization or past myocardial infarction. Pa- seeking all risk factors of atherosclerosis in the pres- tients with a presence of mild valvular disease and ence of ED, as this could be the first sign of this left ventricular dysfunction (NYHA functional class diffuse vascular disorder. The risk categories of Prince- I) also belonged to this category. ton guidelines are listed in Table 11.6. • Patients at indeterminate risk would require further The overall analysis of all randomized controlled testing or evaluation before resuming sexual activ- trials, in which treatment with PDE5 inhibitors was ity. These patients had more than three cardiac risk compared to controls in diabetic patients, demonstrated factors, moderate, stable angina, a history of my- the improvement of ED in diabetic men without re- ocardial infarction of more than 2 weeks but less porting any increased mortality or high frequency of than 6 weeks, a heart failure (NYHA class II) or cardiovascular adverse events. The use of PDE5 in- non-cardiac sequelae of atherosclerotic diseases. hibitors in the presence of oral nitrates is absolutely contraindicated in diabetic men, as in non-diabetic sub- • The high-risk category consisted of patients whose jects. cardiac conditions were sufficiently severe and/or unstable that sexual activity might pose a significant Sildenafil was the first PDE5 inhibitor to risk. These conditions included: unstable or refrac- Sildenafil be developed for the treatment of ED. Sildenafil is tory angina pectoris, uncontrolled hypertension, con- rapidly absorbed following oral administration, and gestive heart failure (NYHA class III or IV), recent has an onset of action within 25–60 min [47]. The myocardial infarction (<2 weeks), high-risk, malig- plasma half life is approximately 4 h. This action nant arrhythmias, hypertrophic obstructive and other profile provides a spontaneity for patients, since in cardiomyopathies, and moderate to severe valvular most of the cases successful intercourse will occur disease. as early as 36 min. The first study in 268 men with Following the categorization of patients, an algorithm both types of diabetes was conducted in 1999, thus should be followed to create the individualized risk establishing the efficacy and safety of sildenafil in this 156 CH 11 ERECTILE DYSFUNCTION population [47]. The mean scores of the self- that diabetic patients with more severe endothelial administered International Index of Erectile Function dysfunction might derive a greater benefit from the reflected significant improvements, and successful NO-potentiating effect of tadalafil. attempts at sexual intercourse were fourfold higher A retrospective analysis of the data from 637 dia- in comparison to placebo. Erections were improved betic men from 12 placebo-controlled trials provided by 56% in the sildenafil group, and by 10% in the several conclusions on the effect of tadalafil on ED in placebo group, and the frequency of cardiovascular diabetes [53]. First, the baseline severity of ED was events in sildenafil-treated men was comparable to more pronounced in diabetic men than in controls. that in controls. Despite the more severe ED, tadalafil efficiently im- The first study in type 2 diabetic men included pa- proved all parameters recorded in questionnaires and tients with a disease duration of at least 2 years, with diaries. This improvement was slightly lower in dia- similar efficacy data to the previously described study betic men than in non-diabetic individuals, as a similar [48]. These investigations showed a more pronounced result was found in studies with the two further PDE5 effect on erection than the previous trial on mixed di- inhibitors. Neither glycaemic control nor the applied abetic men: sildenafil was better in 65% of treated antidiabetic therapy influenced the effect of tadalafil. patients and in 11% of placebo-receiving subjects. In The sustained efficacy for up to 36 h was evident in dia- addition, the treatment was well tolerated and inde- betic men treated with tadalafil. The adverse event pro- pendent of the degree of glycaemia and presence of file observed in diabetic patients was similar to that in neuropathy or vascular disease; this suggested would non-diabetic patients. An absence of the more frequent also be effective in diabetic patients with complica- serious events of myocardial ischaemia in diabetic pa- tions. A study in type 1 diabetic patients [49] also tients treated with tadalafil strengthened the previous established better mean scores of questionnaires and observation that PDE5 inhibition is safe if patients are an improved erection in the sildenafil group, by 66%. categorized by their cardiovascular risk before the in- The rate of adverse events was low and the sildenafil dication. Recently, two regimens of tadalafil dosing was effective, irrespective of the severity of the ED. A were compared in 752 diabetic men [54], namely an further study [50] found that men with both types of di- on-demand treatment and a regular treatment (three abetes had a lower response rate to sildenafil than had times weekly). Although the efficacy measures and been previously reported in large, non-diabetic popula- treatment satisfaction data reflected therapeutic success tions; moreover, a higher rate of cardiovascular adverse for both dosing schedules, the treatment preferences events was found than in other trials (7% versus 0%; differed, this being 57.2% for on-demand regimen in comparison with placebo). compared to 42.8% for three-times daily regimen.

Tadalafil The mean half-life of tadalafil is 17.5 h, Vardenafil Single doses of vardenafil (10–40 mg) are and a dose of 10 or 20 mg is well tolerated for up rapidly absorbed following oral administration, with to 36 h. This duration permits once-a-day dosing, such maximum plasma concentrations being achieved in that patients are highly compliant with this regimen some men within 15 min [55]. A high-fat meal, how- [51]. Up to once daily, on-demand dosing was tested ever, reduces the rate of absorption (similarly with for 12 weeks in 191 type 1 and type 2 diabetic men in sildenafil), such that the time to maximum plasma con- 2002 [52]. Therapy with tadalafil in this group consis- centration is increased to 1 hour. A prospective study tently enhanced erectile function, and also significantly in which a large number of diabetic patients (n = 452) improved the patients’ ability to achieve and maintain was enrolled analyzed the efficacy of the fixed dos- an erection. The percentage of men reporting improved ing (10 or 20 mg) of vardenafil on ED in Canada and erections in this study was 56% (10 mg/day) and 64% the US [56]. A dose-dependent, clinically meaningful, (20 mg/day). Similar increases in the proportions of statistically significant improvement was found in all positive responses to questionnaires were proven, and three primary efficacy measures of erectile function. tadalafil was seen to be well tolerated in men with The success rates were independent of baseline ED diabetes and ED, regardless of the HbA1c level. In- severity, the level of glycaemic control, and irrespec- terestingly, in this study, those patients taking con- tive of whether patients had type 1 or type 2 diabetes. A comitant antihypertensive medications had greater im- study in Germany was designed to evaluate the efficacy provements in erectile function with tadalafil at 20 mg of flexible dosing (5–20 mg) of tadalafil in diabetic than those not taking antihypertensives. This suggested men [57]. This method of treatment also achieved a 11.9 TREATMENT OF ERECTILE DYSFUNCTION IN DIABETIC PATIENTS 157 significant improvement in several diary questions of transurethral or intraurethral and topical administration the Sexual Encounter Profile. A similar beneficial ef- of vasoactive drugs. fect was found in patients with poor, moderate or good glycaemic control. In a study on 778 Japanese diabetic Intracavernosal therapy men with ED, it became clear that although both 10 For the intracavernosal procedure, alprostadil or a mix- and 20 mg vardenafil doses were effective in improv- ture of drugs are administered. Alprostadil has the ing erectile function, the 20 mg dose demonstrated a same effect with prostaglandin E1; thus the α1-blocking superior efficacy compared to 10 mg, which suggested properties mediated through a membrane receptor re- an incremental clinical benefit of using the higher dose lax the cavernous and arteriolar smooth muscle [60]. in this difficult-to-treat population [58]. Alprostadil produces full erections at doses as low as As a conclusion of the experience with PDE5 in- 2.5 μg. An open-label, flexible dose-escalating study hibitors in diabetic men, it can be stated that these involving 336 men was designed to evaluate the ef- drugs have improved erectile function for all efficacy ficacy of this intracavernosal treatment in diabetes variables, to a significantly greater extent than the [61]. All men were fully trained in the self-injection placebo control [59]. Moreover, this treatment is well technique before entry into the home phase, during tolerated and safe, when following the considerations which a satisfactory erectile response was achieved of the cardiovascular risk stratifications of the Prince- after 99% of injections, while the median alprostadil ton guidelines. As the higher dose results in a greater dose remained unchanged. Unfortunately, however, response in these patients, the higher available dose 24% of the study group reported penile pain and/or should be used in these cases. The efficacy is indepen- injection-site pain. The concept of the administration of dent of the type of the metabolic disease or the gly- multiple vasoactive drugs is based on the possible syn- caemic control. The differences in clinical applications ergistic effect derived from the different mechanisms between PDE5 inhibitors in diabetic men – other than of action that produce the erectogenic effects. In addi- dosage regimens and their efficacy in patients with se- tion, a lower frequency of adverse events is expected vere retinopathy or autonomic neuropathy – will hope- as it is not necessary to administer the total dose of the fully be proven by further studies. The advantages and separate drugs. The most frequently used preparation disadvantages of PDE5 inhibitors in diabetic patients is a three-drug mixture composed of papaverine, phen- are summarized in Table 11.7. tolamine and alprostadil [62]. Although the response rate to these multidrug mixtures is high, most are not 11.9.2 Second-line therapy approved by national health care authorities [60]. Second-line therapy is considered only in those con- Intraurethral therapy ditions when the first-line therapy is ineffective or The efficacy of intraurethral treatment is explained by a is contraindicated, and involves the intracavernosal, transfer of drug from the urethra directly into the corpora cavernosa. Alprostadil is applied intraurethrally Table 11.7 Advantages and disadvantages of PDE5 via a special applicator, the Medicated Urethral System inhibitors in diabetic patients. for Erection (MUSE). The results demonstrated a suc- Advantages Disadvantages cessful penetration which occurred at least once during • Oral administration • Contraindicated in males a trial, in 65–70% of patients with MUSE compared to 10–20% with placebo [63]. Although MUSE is more • Definitely effective in receiving nitrate treat- comfortable for most patients than intracavernosal in- prospective studies ment • jections, alprostadil is significantly less effective when • Influence on a main ae- Careful risk assessment is required applied in this way [64]. Both treatment modalities re- tiological factor quire that the patients be trained at the doctor’s surgery • • Efficacy independent of Dose regimen is different before home administration. glycaemic state or type from the general popula- of disease tion Topical therapy • The efficacy in patients with severe diabetic com- Topical drug delivery represents a simple, reversible, plications is not proven non-invasive, spontaneous second-line treatment option for erectile dysfunction, but the efficacy of this 158 CH 11 ERECTILE DYSFUNCTION application must be facilitated by an enhanced degree References of skin and tunica permeation [65]. Nitroglycerine, papaverine, minoxidil and alprostadil are administered in 1. NIH Consensus Development Panel on Impotence. this way. (1993) JAMA, 270: 83–90. 2. Solomon H, Man JW, Wierzbicki AS and Jackson G. Relation of erectile dysfunction to angiographic 11.9.3 Third-line therapy coronary artery disease. Am J Cardiol 2003; 91 (2): 230–1. Third-line therapy for ED is applied when there is 3. O’Kane PD and Jackson G. Erectile dysfunction: is an intolerance or lack of response to other treatment there silent obstructive coronary artery disease? Int J modalities. Two types of rare invasive therapeutic in- Clin Pract 2001; 55 (3): 219–20. tervention are recognized in this category among the 4. Feldman HA, Goldstein I, Hatzichristou DG, Krane general population of men with ED, namely the im- RJ and McKinlay JB. Impotence and its medical and plantation of a penile prosthesis and penile revascular- psychosocial correlates: results of the Massachusetts Male Aging Study. JUrol1994; 151: 54–61. ization [62]. 5. Fonseca V and Jawa A. Endothelial and erectile A penile prosthesis is chosen when an irrepara- dysfunction, diabetes mellitus, and the metabolic bly damaged erectile tissue is present as a result syndrome: common pathways and treatments? The of a permanent extensive damage of the cavernosal American Journal of Cardiology 2005; 96 (Suppl.): smooth muscle with replacement by fibrous tissue. In 13M–18M. this case, the intracavernosal spongy tissue becomes 6. Romeo JH, Seftel AD, Madhun ZT and Aron DC. Sex- non-compliant and non-responsive to pharmacological ual function in men with diabetes type 2: association stimulation. While this severe organic manifestation is with glycemic control. JUrol2000; 163: 788–91. not directly related to diabetes, it can develop as a 7. McCulloch DK, Campbell IW, Wu FC, Prescott RJ result of an intercurrent condition (irreversible vascu- and Clarke BF. The prevalence of diabetic impotence. lar damage, chronic renal disease, priapism, Peyronie’s Diabetologia 1980; 18: 279–83. disease, penile trauma). Both, malleable and inflatable 8. Klein R, Klein BE, Lee KE, Moss SE and Cruickshanks KJ. Prevalence of self-reported erectile dysfunction in penile prostheses are available, and a high efficacy has people with long-term IDDM. Diabetes Care 1996; 19: been reported. Preoperative psychiatric screening must 135–41. be performed, however, in order to exclude those pa- 9. Tesfaye S, Stevens LK, Stephenson JM, Fuller JH, tients who will not tolerate the device. Plater M, Ionescu-Tirgoviste C, Nuber A, Pozza G, Surgery for penile revascularization has proved not Ward JD and the EURODIAB IDDM Study Group. to be efficient among different special populations Prevalence of diabetic peripheral neuropathy and its of patients, and thus is currently contraindicated in relation to glycaemic control and potential risk fac- diabetes. tors: the EURODIAB IDDM Complications Study. Di- abetologia 1996; 39: 1377–84. 10. Sun P, Cameron A, Seftel A, Shabsigh R, Niederberger 11.10 Conclusions C and Guay A. Erectile dysfunction – an observable marker of diabetes mellitus? A large national epidemi- As diabetes represents a very serious risk factor for ological study I. Urology 2006; 176: 1081–5. ED, all therapeutic options should be considered at 11. Barrett-Connor E. Heart disease risk factors predict the initiation of treatment. These most important steps erectile dysfunction 25 years later (The Rancho to achieve an effective therapy include a careful ex- Bernardo Study). Am J Cardiol 2005; 96 (Suppl.): ploration of the patient’s complaints, the suggestion 3M–7M. 12. Haczynski J, Lew-Starowicz Z, Darewicz B, Krajka of a lifestyle change, and the administration of PDE5 K, Piotrowicz R and Ciesielka B. The prevalence of inhibitors, assuming a lack of contraindication. One erectile dysfunction in men visiting outpatient clinics. of the most important elements when considering the International Journal of Impotence Research 2006; 18: therapy of ED in diabetic men is to design the treat- 359–63. ment individually, with particular attention being paid 13. Bacon CG, Mittleman MA, Kawachi I, Giovannucci to the patient’s medical, psychiatric and social condi- E, Glasser DB and Rimm EB. Sexual function in men tions. older than 50 years of age: results from the health 11.10 CONCLUSIONS 159

professionals follow-up study. AnnInternMed2003; 26. Montorsi P, Ravagnani PM, Galli S, Salonia A, Brig- 139: 161–8. anti A, Werba JP and Montorsi F. Association be- 14. Walczak MK, Lokhandwala N, Hodge MB and Guay tween erectile dysfunction and coronary artery disease: AT. Prevalence of cardiovascular risk factors in erectile matching the right target with the right test in the right dysfunction. J Gend Specif Med 2002; 5: 19–24. patient. Eur Urol 2006; 50 (4): 721–31. 15. Chung WS, Sohn JH and Park YY. Is obesity an 27. Solomon H, Man JW, Wierzbicki AS and Jackson G. underlying factor in erectile dysfunction? Eur Urol Relation of erectile dysfunction to angiographic artery 1999; 36: 68–70. disease. Am J Cardiol 2002; 91: 230–1. 16. Sun P and Swindle R. Are men with erectile dysfunc- 28. Bortolotti A, Parazzini F, Colli E and Landoni M. tion more likely to have hypertension than men with- The epidemiology of erectile dysfunction and its risk out erectile dysfunction? A naturalistic national cohort factors. Int J Androl 1997; 20: 323–34. study. JUrol2005; 174: 244–8. 29. Gazzarusso C, Giordanetti S, De Amici E, et al.Re- 17. Billups KL, Bank AJ, Padma-Nathan H, Katz S and lationship between erectile dysfunction and silent my- Williams R. Erectile dysfunction is a marker for cardio- ocardial ischemia in apparently uncomplicated type 2 vascular disease: results of the Minority Health Institute diabetic patients. Circulation 2004; 110: 22–6. Expert Advisory Panel. JSexMed2005; 2: 40–52. 30. Virag R and Bouilly P. Is impotence an arterial disease? 18. Burchardt M, Burchardt T, Anastasiadis AG, Kiss AJ, A study of arterial risk factors in 440 impotence men. Shabsigh A, de La Taille A, Pawar RV, Baer L and Lancet 1985; 322: 181–4. Shabsigh R. Erectile dysfunction is a marker for cardio- 31. Montorsi P, Ravagnani PM, Galli S, Rotatori F, Brig- vascular complications and psychological functioning anti A, Salonia A, Rigatti P and Montorsi F. The artery in men with hypertension. Int J Impot Res 2001; 13: size hypothesis: a macrovascular link between erectile 276–81. dysfunction and coronary artery disease. Am J Cardiol 19. Nicolosi A, Glasser DB, Brock G, Laumann E and 2005; 96 (Suppl.): 19M–23M. Gingell C. Diabetes and sexual function in older adults: 32. Vileikyte L, Leventhal H, Gonzalez JS, Peyrot M, Ru- results of an international survey. Br J Diabetes Vasc bin R, Ulbrecht J, Garrow A, Waterman C, Cavanagh P Dis. 2002; 2: 336–9. and Boulton AJM. Diabetic peripheral neuropathy and 20. Jiaan DB, Seftel AD, Fogarty J, Hampel N, Cruz W, depressive symptoms: the association revisited. Dia- Pomerantz J, Zuik M and Monnier VM. Age-related betic Care 2005; 28: 2378–83. increase in an advanced glycation end product in penile 33. Haro JM, Beardsworth A, Casariego J, Gavart S, tissue. World J Urol 1995; 13: 369–75. Hatzichristou D, Martin-Morales A, Schmitt H, Mirone 21. Jevtich MJ, Edson M, Jarman WD and Herrera HH. V, Needs N, Riley A, Varanese L, von Keitz A and Vascular factor in erectile failure among diabetics. Kontodimas S. Treatment-seeking behavior of erectile Urology 1982; 19: 163–8. dysfunction patients in Europe: Results of the Erectile 22. Seftel AD, Vaziri ND, Ni Z, Razmjouei K, Fogarty Dysfunction Observational Study. JSexMed2006; 3 J, Hampel N, Polak J, Wang RZ, Ferguson K, Block (3): 530–40. C and Haas C. Advanced glycation end products in 34. Corona G, Mannucci E, Mansani R, Petrone L, Bar- human penis: elevation in diabetic tissue, site of de- tolini M, Giommi R, Forti G and Maggi M. Organic, position, and possible effect through iNOS or eNOS. relational and psychological factors in erectile dysfunc- Urology 1997; 50: 1016–26. tion in men with diabetes mellitus. Eur Urol. 2004 46 23. Burchardt T, Burchardt M, Karden J, Buttyan R, Shab- (2): 222–8. sigh A, de la Taille A, Ng PY, Anastasiadis AG and 35. Rosen RC, Friedman M and Kostis JB. Lifestyle man- Shabsigh R. Reduction of endothelial and smooth mus- agement of erectile dysfunction: the role of cardiovas- cle density in the corpora cavernosa of the streptozo- cular and concomitant risk factors. Am J Cardiol 2005; tocin induced diabetic rat. JUrol2000; 164: 1807–11. 96 (12B): 76M–79M. 24. Vernet D, Cai L, Garban H, Babbitt ML, Murray 36. Jardin A, Wagner G and Khoury S. (2000) Recommen- FT, Rajfer J and Gonzalez-Cadavid NF. Reduction of dations of the 1st international consultation on erectile penile nitric oxide synthase in diabetic BB/WORdp dysfunction. In: A Jardin, G Wagner, S Khoury, F Giu- (type I) and BBZ/WORdp (type II) rats with erectile liano, H Padma-Nathan and R Rosen (eds). Erectile dysfunction. Endocrinology 1995; 136: 5709–17. Dysfunction. Plymbridge Distributors Ltd.: Plymouth, 25. Spark RF, White RA and Connolly PB. Impotence UK, pp. 711–26. is not always psychogenic: newer insights into 37. Richardson D and Vinik A. Etiology and treatment of hypothalamic-pituitary-gonadal dysfunction. JAMA erectile failure in diabetes mellitus. Cur Diab Rep 2002; 1980; 243: 750–5. 2: 501–9. 160 CH 11 ERECTILE DYSFUNCTION

38. Francis ME, Kusek JW, Nyberg LM and Eggers PW. 49. Stuckey BG, Jadzinsky MN, Murphy LJ, Montorsi F, The contribution of common medical conditions and Kadioglu A, Fraige F, Manzano P and Deerochana- drug exposures to erectile dysfunction in adult males. wong C. Sildenafil citrate for treatment of erectile JUrol2007 178 (2): 591–6. dysfunction in men with type 1 diabetes: results of a 39. Lundberg PO and Biriell C. Sexual dysfunction as a randomized controlled trial. Diabetes Care 2003; 26 suspected drug reaction reported to WHO Collaborat- (2): 279–84. ing Centre for International Drug Monitoring. Therapie 50. Safarinejad MR. Oral sildenafil in the treatment of 1993; 48 (5): 457–9. erectile dysfunction in diabetic men: a randomized 40. Rosen R. Psychogenic erectile dysfunction: classifica- double-blind and placebo-controlled study. J Diabetes tion and management. Urol Clin North Am. 2001; 28: Complications 2004; 18 (4): 205–10. 269–78. 51. Porst H, Giuliano F, Glina S, Ralph D, Casabe AR, 41. Shabsigh R, Rajfer J, Aversa A, Traish AM, Yassin Elion-Mboussa A, Shen W and Whitaker JS. Evalu- A, Kalinchenko SY and Buvat J. The evolving role of ation of the efficacy and safety of once-a-day dosing testosterone in the treatment of erectile dysfunction. Int of tadalafil 5 mg and 10 mg in the treatment of erec- J Clin Pract. 2006; 60 (9): 1087–92. tile dysfunction: results of a multicenter, randomized, 42. American Association of Clinical Endocrinologists double-blind, placebo-controlled trial. Eur Urol 2006, medical guidelines for clinical practice for the 50 (2): 351–9. evaluation and treatment of hypogonadism in adult 52. Saenz de Tejada I, Anglin G, Knight JR and Emmick male patients – 2002 update. Endocr Pract. 2002; 8 JT. Effects of tadalafil on erectile dysfunction in men (6): 439–56. with diabetes. Diabetes Care 2002; 25 (12): 2159–64. 43. Isidori AM, Giannetta E, Gianfrilli D, Greco EA, 53. Fonseca V, Seftel A, Denne J and Fredlund P. Impact Bonifacio V, Aversa A, Isidori A, Fabbri A and Lenzi of diabetes mellitus on the severity of erectile dysfunc- A. Effects of testosterone on sexual function in men: tion and response to treatment: analysis of data from results of a meta-analysis. Clin Endocrinol (Oxf). 2005; tadalafil clinical trials. Diabetologia 2004, 47 (11): 63 (4): 381–94. 1914–23. 44. Schwarz ER, Kapur V, Rodriguez J, Rastogi S and 54. Buvat J, van Ahlen H, Schmitt H, Chan M, Kuepfer C Rosanio S. The effects of chronic phosphodiesterase-5 and Varanese L. Efficacy and safety of two dosing reg- inhibitor use on different organ systems. Int J Impot imens of tadalafil and patterns of sexual activity in men Res 2007; 19 (2): 139–48. with diabetes mellitus and erectile dysfunction: Sched- 45. DeBusk R, Drory Y, Goldstein I, Jackson G, Kaul S, uled use vs. on-demand regimen evaluation (SURE) Kimmel SE, Kostis JB, Kloner RA, Lakin M, Meston study in 14 European countries. JSexMed2006; 3 (3): CM, Mittleman M, Muller JE, Padma-Nathan H, Rosen RC, Stein RA and Zusman R. Management of sexual 512–20. dysfunction in patients with cardiovascular disease: 55. Montorsi F, Salonia A, Briganti A, Barbieri L, Zanni recommendations of The Princeton Consensus Panel. G, Suardi N, Cestari A, Montorsi P and Rigatti P. Am J Cardiol 2000; 86 (2): 175–81. Vardenafil for the treatment of erectile dysfunction: 46. Kostis JB, Jackson G, Rosen R, Barrett-Connor E, a critical review of the literature based on personal Billups K, Burnett AL, Carson C, III, Cheitlin M, clinical experience. Eur Urol 2005; 47 (5): 612–21. Debusk R, Fonseca V, Ganz P, Goldstein I, Guay 56. Goldstein I, Young JM, Fischer J, Bangerter K, A, Hatzichristou D, Hollander JE, Hutter A, Katz Segerson T and Taylor T. Vardenafil, a new phospho- S, Kloner RA, Mittleman M, Montorsi F, Montorsi diesterase type 5 inhibitor in the treatment of erectile P, Nehra A, Sadovsky R and Shabsigh R. Sexual dysfunction in men with diabetes: a multicenter dysfunction and cardiac risk (the Second Princeton double-blind placebo-controlled fixed-dose study. Consensus Conference). Am J Cardiol 2005; 96 (2): Diabetes Care 2003; 26: 777–83. 313–21. 57. Ziegler D, Merfort F, van Ahlen H, Yassin A, Reblin T 47. Rendell MS, Rajfer J, Wicker PA and Smith MD. Silde- and Neureither M. Efficacy and safety of flexible-dose nafil for treatment of erectile dysfunction in men with vardenafil in men with type 1 diabetes and erectile diabetes: a randomized controlled trial. The Sildenafil dysfunction. JSexMed2006; 3 (5): 883–91. Diabetes Study Group. JAMA 1999; 281 (5): 421–6. 58. Ishii N, Nagao K, Fujikawa K, Tachibana T, Iwamoto 48. Boulton AJ, Selam JL, Sweeney M and Ziegler D. Y and Kamidono S. Vardenafil 20-mg demonstrated su- Sildenafil citrate for the treatment of erectile dysfunc- perior efficacy to 10-mg in Japanese men with diabetes tion in men with type II diabetes mellitus. Diabetologia mellitus suffering from erectile dysfunction. Int J Urol 2001; 44(10): 1296–301. 2006; 3 (8): 1066–72. 11.10 CONCLUSIONS 161

59. Vickers M and Satyanarayana R. Phosphodiesterase 63. Ekman P, Sjogren L, Englund G and Persson BE. type 5 inhibitors for the treatment of erectile dysfunc- Optimizing the therapeutic approach of transurethral tion in patients with diabetes mellitus. Int J Impot Res alprostadil. Br J Urol 2000; 86: 68–74. 2002, 14: 466–71. 64. Shabsigh R, Padma-Nathan H, Gittleman M, McMur- 60. Montorsi F, Salonia A, Zanoni M, Pompa P, Cestari A, ray J, Kaufman J and Goldstein I. Intracavernous al- Guazzoni G, Barbieri L and Rigatti P. Current status of prostadil alfadex is more efficacious, better tolerated, local penile therapy. Int J Impot Res 2002; 14 (Suppl. and preferred over intraurethral alprostadil plus op- 1): 70–81. tional ACTIS: a comparative, randomised, crossover, 61. Heaton JP, Lording D, Liu SN, Litonjua AD, Guangwei multicenter study. Urology 2000; 55: 109–13. L, Kim SC, Kim JJ, Zhi-Zhou S, Israr D, Niazi D, 65. Goldstein I, Payton TR and Scechter PJ A double Rajatanavin R, Suyono S, Benard F, Casey R, Brock G -blind, placebo-controlled, efficacy and safety study of and Belanger A. Intracavernosal alprostadil is effective topical gel formulation of 1% alprostadil (Topiglan) for for the treatment of erectile dysfunction in diabetic the in-office treatment of erectile dysfunction. Urology men. Int J Impot Res 2001; 13 (6): 317–21. 2001; 57: 301–5. 62. Vickers MA and Wright EA. Erectile dysfunction in the patient with diabetes mellitus. Am J Manag Care 2004; 10 (Suppl. 1): 3–11. SECTION IV Treatment and Care Issues

Raffaele Marfella and Giuseppe Paolisso Department of Geriatric and Metabolic Disease, Second University Naples, Piazza Miraglia, Naples, Italy

Key messages 12.1 The definition of metabolic syndrome • Metabolic syndrome is a potent indicator of future risk of type 2 diabetes and concomitant The ‘metabolic syndrome’, introduced by Reaven increased potential for cardiovascular morbidity in 1988 [1], has been characterized by the si- and mortality. multaneous presence of hypertension, a degree of • Worldwide, 3.2 million deaths are attributable glucose intolerance, high triglyceride levels and to diabetes every year – that is 8700 deaths each low high-density lipoprotein (HDL) concentrations. day, or six deaths each minute. At least one in The basic abnormalities underlying these alterations 10 deaths among adults aged 35–64 years is have been indentified in resistance of insulin to attributable to diabetes. mediate glucose utilization. In this context, the • Metabolic syndrome is a common condition, term ‘insulin resistance syndrome’ has often been and its frequency is rising dramatically world- proposed to define this aggregation of risk factors. wide. At least 45% of people aged >60 years The syndrome has also been referred to as ‘metabolic have metabolic syndrome. syndrome’, the ‘plurimetabolic syndrome’ and the • The global increase in metabolic syndrome will ‘deadly quartet’. Recently, the National Cholesterol occur due to population ageing and growth, and Education Program’s Adult Treatment Panel III report to increasing trends towards obesity, unhealthy (NCEP ATP III) has underlined the central role of diets and sedentary lifestyles. • this syndrome in the prevention of cardiovascular A full and healthy life is possible with metabolic disease (CVD) [2], Defining this clustering of syndrome. With good management, many of metabolic risk factors as “metabolic syndrome”. its components can be prevented or delayed. This definition suppose that insulin resistance is the Medication may be used to control blood primary cause of associated risk factors. However, glucose, blood pressure and blood lipids. In several other metabolic abnormalities have been high-risk subjects, lifestyle changes are much associated with the syndrome, including obesity more effective than drugs. (notably abdominal obesity), high apolipoprotein • Optimal health care can substantially reduce the (apo) B levels, reduced low-density lipoprotein (LDL) risk of developing cardiovascular diseases. levels, and also abnormalities in fibrinolysis and coagulation [3].

12.1.1 How should metabolic syndrome be Table 12.2 WHO clinical criteria for the metabolic defined clinically? syndrome. Three health authorities have provided practical criteria Insulin resistance, identified by ONE of the following: to identify patients with metabolic syndrome, however, • Type 2 diabetes these criteria differ between organizations [2, 4, 5]. • Impaired fasting glucose The tools suggested by the NCEP ATP III, the World • Impaired glucose tolerance Health Organization (WHO) and the American Associ- • ation of Clinical Endocrinologists (AACE) to identify, Or for those with normal fasting glucose levels (<110 mg/dL), glucose uptake below the lowest quar- clinically the metabolic syndrome are summarized in tile for background population under investigation under Tables 12.1, 12.2 and 12.3, respectively. hyperinsulinaemic, euglycaemic conditions For the NCEP ATP III, when three of the five criteria listed in Table 12.1 are present, a diagnosis of Plus any TWO of the following: metabolic syndrome can be made. As evidenced in • Antihypertensive medication and/or high blood pressure ≥ ≥ Table 12.1, this guideline considers waist circumfer- ( 140 mmHg systolic or 90 mmHg diastolic) ence as a criteria to evaluate the level of adiposity, • Plasma triglycerides ≥150 mg/dL (≥1.7 mmol l−1) while an explicit demonstration of insulin resistance • HDL cholesterol <35 mg/dL (<0.9 mmol l−1)inmenor (IR) is not required. <39 mg/dL (1.0 mmol l−1)inwomen In contrast, the WHO guidelines view IR, as a • BMI >30 kg m−2 and/or waist:hip ratio >0.9 in men, required component for diagnosis. In addition to IR, >0.85 in women two other risk factors are required for a diagnosis of • ≥ μ the metabolic syndrome. Microalbuminuria has been Urinary albumin excretion rate 20 g/ min or albu- ≥ −1 also added to the list as a criterion. min:creatine ratio 30 mg g Finally, the AACE criteria seem to be a compro- Derived from Ref. [4]. mise between the NCEP ATP III and WHO guidelines. Moreover, the number of risk factors required to make the diagnosis of the metabolic syndrome is Table 12.3 AACE clinical criteria for the metabolic not specified and left to clinical judgement. Both, syndrome. the AACE and WHO guidelines have included IR Risk factor Cutpoints for abnormality measurements that are beyond routine clinical assess- components ment. Indeed, glycemic alterations derived from an −2 oral glucose tolerance test are included in the risk Overweight/obesity BMI ≥25 kg m factors useful to identify the metabolic syndrome. Elevated triglycerides ≥150 mg/dL (1.7 mmol l−1) Low HDL choleterol Table 12.1 NCEP ATP III clinical criteria for the Men <40 mg/dL (1.0 mmol l−1) metabolic syndrome. Women <50 mg/dL (1.3 mmol l−1) Risk factors Defining level Elevated blood ≥130/85 mmHg Abdominal obesity, given pressure as waist circumference 2-Hour postglucose >140 mg/dL Men >102 cm (>40 in) challenge Women >88 cm (>35 in) Fasting glucose Between 110 and 126 mg/dL Triglycerides ≥150 mg/dL (≥1.7 mmol l−1) Other risk factors Family history of type 2 HDL cholesterol diabetes, hypertension, or CVD, polycystic ovary Men <40 mg/dL (<1.0 mmol l−1) syndrome, sedentary lifestyle, Women <50 mg/dL (<1.3 mmol l−1) advancing age, ethnic groups Blood pressure ≥130/≥85 mmHg having high risk for type 2 Fasting glucose ≥110 mg/dL (≥6.1 mmol l−1) diabetes or CVD

Derived from Ref. [2]. Derived from Ref. [5]. 12.2 PATHOGENESIS OF THE METABOLIC SYNDROME 167 • Category 1 = obesity and disorders of adipose tissue. • Category 2 = insulin resistance. • Category 3 = a constellation of independent factors (e.g. molecules of hepatic, vascular and immunolog-

Prevalence (%) ical origin) that mediate specific components of the Age (years) metabolic syndrome [8]. The visceral obesity may be responsible for the clustering of risk factors. This hypothesis is supported by Figure 12.1 Age-specific prevalence of the metabolic the fact that visceral obesity is strongly associated with syndrome among 8814 US adults aged at least 20 years, all cardiovascular risk factors [9, 10]. Therefore, it is by gender. National Health and Nutrition Examination possible to define the metabolic syndrome as a cluster Survey III, 1988–1994. Reproduced from Ref. [6]. of the metabolic complications associated with obesity (more specifically visceral obesirty). Adipose tissue is These definitions of the metabolic syndrome have now recognized as an endocrine organ that secretes nu- allowed to define prevalence of this condition in the merous proteins which exert a variety of effects [11]. worldwide population. Overall, the unadjusted preva- Indeed, hyperplasia and hypertrophy of adipocytes (as seen in obesity) leads to an increased production of lence of the metabolic syndrome was approximately leptin, resistin, acylation-stimulating protein and many 22% in this US adult population. However, the preva- other proteins, as well as hypertrophic adipocytes pro- lence increases with age, both in men and women, duce various proinflammatrory cytokines implicate in to reach almost 45% for subjects aged >60 years the IR and atherosclerotic processs such as tumor (Figure 12.1) [6]. Moreover, considering the epidemic necrosis factor (TNF-alpha) and interleukin (IL) 6, of obesity [7], these numbers will increase in the near and also a decreased production of adiponectin that future. The metabolic syndrome is a rapidly grow- have a protective effects on vascular health. Through ing threat to public health, and a major challenge that these mechanisms, it is clear that obesity plays a cen- physicians and public health agencies must face. tral role in the pathogenesis of the metabolic syndrome. Moreover, this role is also accentuated by the 12.2 Pathogenesis of the metabolic observations that weight loss reducing proinflamma- tory citokines and leptin improves insulin sensitivity syndrome [12]. Indeed, there is substantial evidence that weight loss – particularly the mobilization of visceral adi- The interactions of many susceptibility genes and many pose tissue – leads to simultaneous improvements of environmental exposures play a pivotal role in the the metabolic profile. Taken together, these arguments pathogenesis of the metabolic syndrome. These con- place (visceral) obesity at the heart of the metabolic siderations are supported by the observations that each syndrome. Despite these facts, disagreement persists component of the metabolic syndrome is regulated as to whether IR or abdominal obesity is the primary through both genetic and environmental factors. In- contributor to the metabolic syndrome. It is true that deed, physical inactivity, combined with an athero- there is a broad range of insulin sensitivity at any genic diet (rich in saturated fat, trans fatty acids and re- given level of body fat, and a large spectrum of obe- fined sugars), are responsible for the rising prevalence sity at any given level of insulin sensitivity [13, 14]. of the metabolic syndrome. These lifestyle variables, However, this also means that not all insulin-resistant determining obesity as well as insulin resistance, alter individuals are overweight, nor all overweight individ- the metabolic homeostasis and lead to the multiplex uals are insulin-resistant. Some investigators place a risk factors. greater priority on IR by arguing that insulin resis- The National Heart, Lung and Blood Institute tance/hyperinsulinaemic individuals, with or without (NHLBI), in collaboration with the American Heart obesity, are more likely to display the abnormalities Association (AHA), have identified three potential of the metabolic syndrome [15]. They believe that IR aetiological categories of the metabolic syndrome: or hyperinsulinaemia directly causes other metabolic 168 CH 12 METABOLIC RISK FACTORS, OBESITY AND CARDIOMETABOLIC SYNDROME risk factors. Finally, one point of agreement is that into clinical practice, and many geriatricians claim that IR generally increases with body fat content [14]. It the administration of any available medical treatments became also evident that central obesity is highly cor- is still conditioned to a previous diagnosis of specific related with IR, and consequently central obesity may diseases. Therefore, only new clinical trials aimed at be a surrogate for IR. It is also clear that both factors treating metabolic syndrome components in older per- can play an independent role in the syndrome, giving sons will be required to focus on significant differences their independent effects on cardiovascular risk factors when treating older, middle-aged and younger persons and CVD. Finally, one point is certain, however – that with metabolic syndrome. the rising prevalence of metabolic syndrome is a direct consequence of obesity epidemia, which in turn is 12.4 Insulin resistance and driven by changeable factors such as high-calorie diets and a sedentary lifestyle. dyslipidaemia

Increased plasma triglycerides, reduced LDL-choles- 12.3 Metabolic syndrome in older terol levels and reduced HDL-cholesterol levels con- persons stitute the atherogenic dyslipidaemia of metabolic syndrome. In particular, IR is associated with: (i) hy- As noted above, the key player in the metabolic syn- pertriglyceridaemia and a subsequent increase in very drome is IR. Moreover, due to the fact that the aging low density lipoproteins (VLDL); (ii) small and dense process is significantly associated with an increase in LDL; and (iii) lower HDL levels. Visceral fat seems IR per se, older persons with metabolic syndrome must to have a major role and, indeed, the quantity of vis- be recognized as a delicate group of individuals with ceral fat tissue [measured using computed tomography an increased risk of developing type 2 diabetes mel- (CT)], waist circumference and IR are all significantly litus and CVD. The aging process is associated with linked one to another [21, 23]. Visceral fat tissue and impaired glucose handling, mainly due to a decline in IR are known to be at the basis of an increased hep- insulin action [16–19]. There is strong evidence that atic production of triglycerides and VLDL, as well an increased resistance to insulin action is one of the as a reduced intravascular breakdown of VLDL and main components of diminished homeostatic glucose chylomicrons. Under physiological conditions, insulin regulation in older persons. Insulin-mediated glucose will inhibit lipolysis and promote lipogenesis. Thus, by uptake (measured using a glucose clamp technique) promoting the flow of intermediates through glycoly- was shown to decline progressively with aging [20]. sis, insulin will promote formation of the α-glycerol The unifying hypothesis described by Barbieri et al. phosphate and fatty acids necessary for triglyceride [20], explaining the relationship between IR and age, formation. Insulin then, stimulates fatty acid synthase, encompasses four main pathways: thus leading to an increased fatty acid synthesis. At the same time, in the adipose tissue cell, insulin • Anthropometric changes (an increase in fat mass inhibits the breakdown of triglycerides by inhibiting with a parallel decline in fat free mass). the activity of a hormone-sensitive lipase, the enzyme • Environmental changes (diet habits and reduced required for triglyceride breakdown. This enzyme is physical activity). also found on the luminal surface of the endothelium of capillaries in both adipose and skeletal muscle tissues. • Neurohormonal variations, which may have an op- However, at the level of the endothelium, insulin in- posite effect of that of insulin at skeletal muscle and creases the activity of lipoprotein lipase, which in turn adipose levels. increases the breakdown of triglycerides in VLDLs as • A rise in oxidative stress. well as chylomicrons; this is an important role in the uptake of free fatty acids (FFAs) from the bloodstream Consequently, any one of the individual components of to adipose tissue. As a result, lipoproteins synthesized the metabolic syndrome may have a greater clinical im- in the liver are taken up by adipose tissue and FFAs are portance when present in an older person. Indeed, the ultimately stored as triglycerides. During an IR state, management of metabolic syndrome in older persons insulin action on adipose tissue is no longer effective in will lead clinicians to adopt precise measures. How- repressing the activity of such lipoprotein lipase, and a ever, this approach has not been completely translated consequent rise occurs in the plasma concentrations of 12.5 INSULIN RESISTANCE AND OBESITY 169 FFAs [24]. On the vascular level, a reduced activity of inversely correlated with their triglyceride content – a lipoprotein lipase results in a slower catabolism of chy- phenomenon due to an increased catabolism of HDL lomicrons and VLDLs, and this is clinically expressed itself [28]. The higher concentrations of triglycerides as hypertriglyceridaemia. Such triglycerides will be se- in HDL is due to increased activity of CETP, which creted by the liver as VLDL in circulation, thereby also transfers triglycerides from VLDLs to HDLs [29]. creating a vicious cycle. Fasting hypertriglyceridaemia Consequently, an increased liver lipase activity (this has been identified as an independent risk factor for is no longer effectively inhibited by insulin) results ischaemic heart disease (IHD) [24]; notably, these au- in a smaller core volume of HDL and the formation thors reported that triglyceride-rich lipoproteins have of smaller and more dense HDL. During this phase, a different atherogenic potential. In addition to the di- apo A-I, which normally is present on the surface rect atherogenic effect of triglyceride-rich lipoproteins, of HDL, becomes detached (this can be detected high triglyceride levels appear to be a marker of a se- by significantly higher urinary concentrations) ries of other potentially atherogenic and prothrombotic [30]. These above changes are especially active changes. during an IR state. Hence, all three altered lipid The second observation seen in the atherogenic pro- profiles – hypertriglyceridaemia, smaller and dense file during IR is the presence of small, dense and LDLs and lower HDL concentrations – constitute the highly atherosclerotic LDLs. The exact mechanisms so-called ‘atherogenic lipoprotein profile’ [30]. involved in the formation of small and dense LDL is not completely understood, although the hepatic lipase (HL) enzyme found on the endoluminal surface 12.5 Insulin resistance and obesity of hepatic sinusoids appears to have an important role in lipoprotein size and density formation. Hepatic li- A further component of the metabolic syndrome is pase acts specifically by hydrolyzing phospholipids and obesity, notably the presence of extensive visceral fat triglycerides found in HDLs and intermediate density tissue. It has been shown in many studies that increased lipoproteins (IDLs), and is also known to hydrolyze visceral fat is a risk factor for age-related diseases phospholipids and triglycerides in LDLs [25]. In par- such as hypertension, type 2 diabetes mellitus, CVD ticular, it has been shown that the more active the HL, and reduced cognitive functioning [31–34]. It is well the greater the release of phospholipids and triglyc- known that aging is associated with a decrease in lean erides, and this results in the formation of smaller and body mass (especially muscle tissue) and a parallel more dense LDLs [26]. Considering that IR is asso- rise in fat mass [35]. There is a slow, progressive re- ciated with a significantly greater activity of HL, this distribution of fat as the intra-abdominal fat tends to alteration might explain why a greater concentration increase while the subcutaneous fat on the limbs tends of smaller and denser LDLs are found in IR indi- to decrease. Data have repeatedly demonstrated that viduals. Furthermore, LDLs in the bloodstream of IR intra-abdominal fat is a major clinical parameter asso- persons have higher concentrations of triglycerides due ciated with IR. Although the mechanisms for the link to the cholesterol ester transfer protein (CETP), an en- between IR and intra-abdominal fat accumulation have zyme that transfers triglycerides from VLDLs to LDLs not been fully elucidated, it has been suggested that [27], thus forming the perfect substrate for highly ac- a high lipolytic response of visceral adipose tissue to tive HL (no longer insulin-inhibited) and the formation catecholamine exposes the liver to high FFA concentra- of smaller and more dense LDLs. Small dense LDLs tions, which are known to play a role in IR (see above). might be more atherogenic than normal dense LDLs Increased adipose tissue present in overweight and due to the fact that they: (i) are more toxic to the obese individuals is no longer considered to be an inert endothelium; (ii) have a greater capability of cross- bystander, but rather as an active endocrine organ ca- ing the endothelial membrane; (iii) have an increased pable of regulating whole-body metabolism and other susceptibility to oxidation; and (v) are more selective vital functions related to inflammation and immune to bound to scavenger receptors on monocyte-derived responses [36–41]. These actions are mediated by a macrophages [27]. number of molecules that are secreted by adipocytes The third altered lipid profile observed during and act in an autocrine, paracrine or endocrine fash- metabolic syndrome is the occurrence of lower ion. Among those identified to date are leptin, adipsin, HDL concentrations. The HDL are smaller and resistin and adiponectin, all of which are believed to more dense during an IR state, and their size is adapt metabolic fluxes to the amount of stored energy 170 CH 12 METABOLIC RISK FACTORS, OBESITY AND CARDIOMETABOLIC SYNDROME [42]. It follows that an understanding of the regulation have been shown to be significantly higher in obese and expression of such adipokines in older individ- humans. uals with different body fat density and distribution, may indicate a new target for preventive measures, es- 12.6 Insulin resistance and arterial pecially when considering cardiovascular functioning. The deregulation of the adipokine network has been hypertension implicated in the etiology of IR and other components of the metabolic syndrome, such as glucose intoler- Arterial hypertension, which is another component of ance, obesity, dyslipidaemia and hypertension [43]. the metabolic syndrome, clusters with many metabolic In addition, there is a growing list of adipokines in- diseases such as obesity, type 2 diabetes mellitus, volved in the control of pro-inflammatory markers (e.g. atherosclerosis and dyslipidaemic states. The associ- TNF-α, IL-6, IL-1p, IL-8, IL-10, transforming growth ation between IR and arterial hypertension was docu- factor-p, nerve growth factor) and of acute-phase re- mented some 50 years ago [53], and numerous studies sponse (plasminogen activator inhibitor-1, haptoglobin, have since been conducted to investigate the associa- serum amyloid A) [44, 45]. The production of these tion in detail. In fact, one study showed specifically proteins by adipose tissue is increased in obesity, and that a strong association existed among hypertension, raised circulating levels of several acute-phase proteins hyperinsulinaemia and reduced glucose tolerance in a (as well as of inflammatory cytokines) has led to the large population of patients (n = 2475) [54]. Another view that obesity, characterized by a chronic low-grade study evaluated the ability of hyperinsulinaemia (as a inflammatory state, is linked to IR and the metabolic surrogate measure of IR) in predicting the development syndrome. An increase in fat tissue, especially in the of coronary heart disease (CHD) and hypertension [55] abdominal area, has also been associated with in- in a healthy population. Here, it was found that 25% of creased plasma levels of diverse pro-inflammatory cy- the population with the highest insulin response (post tokines. In particular, the increase in pro-inflammatory glucose challenge) had a significantly higher increase cytokines or adipokines include 1L-6, resistin, TNF-α in the incidence of hypertension (twofold) or CHD and C-reactive protein (CRP) reflect the overproduc- (threefold). Furthermore, the results were found to be tion by the expanded adipose tissue. This production independent of any differences in age, gender or body supports evidence that monocyte-derived macrophages mass index (BMI). Indeed, these results indicated the reside in adipose tissue, and are at least in part the importance of IR and/or hyperinsulinaemia on vascular source of cytokine production locally and in the sys- disease development over a 15-year time frame. The temic circulation. The magnitude of this vicious cycle mechanisms responsible for the relationship between remains unknown. IR and hypertension are multifactorial [56]. In recent years, much attention has been paid to an First, it is important to note that insulin is a va- anti-inflammatory adipokine, adiponectin, which has sodilator when given intravenously to normal-weight also been shown to be involved in the modulation of IR subjects [57], but has a vasoconstrictor effect in IR, [46] and also to demonstrate anti-atherogeneic effects aged and type 2 diabetes mellitus patients [58]. It is [47]. Some studies have shown that low concentrations widely known that insulin is capable of stimulating of adiponectin are linked to myocardial infarction [48] the production of nitric oxide (NO), thus resulting in and to the progression of subclinical coronary heart dis- an increased blood flow, especially in skeletal mus- ease [48]. Previous studies have shown that the aging cle. Due to the fact that a common pathway exists process itself is also associated with a deregulation of between the insulin-dependent release of NO and the the inflammatory response [49–51]. Such deregulation metabolic actions of the insulin, any altered activity of is defined by presence of high plasma concentrations of such a pathway would result in both decreased vasodi- pro-inflammatory cytokines such as, TNF-α,IL-6and latation and a reduced glucose skeletal muscle uptake. acute phase reactive proteins in older persons [52]. A In fact, these alterations would explain the presence growing body of evidence has also shown that TNF-α, of an increased peripheral blood flow, especially in IL-6 and CRP contribute to age-related IR [50]. In skeletal muscle tissue, following an inadequate release particular, diverse pro-inflammatory cytokine concen- of insulin-dependent NO release. trations have been linked to aging, IR and the metabolic Second, insulin has a direct effect on cardiac mus- syndrome. These cytokines are most likely important cle tissue [59] by increasing cardiac output and rate in controlling the degree of IR, as their serum levels through the activity of the sympathetic nervous system 12.7 POTENTIAL COMPONENTS OF THE METABOLIC SYNDROME IN THE ELDERLY 171 (SNS) [60]. During an IR state, the SNS becomes hy- production from adipose tissue is known to have a peractive with subsequent peripheral vasoconstriction, marked influence on the hypercoagulable state by increased heart rate and hypertension. increasing hepatic fibrinogen production, as well as Third, it is widely known that insulin has an causing endothelial dysfunction towards a hypercoag- anti-natriuretic effect by activating the renin- ulable condition. During IR, the simultaneous increase angiotensin-aldosterone system [57]. Studies have of both tissue factor (TF) and factor VII significantly also highlighted a significant and inverse relationship enhances activation of the coagulation cascade. between IR and sodium diet restriction in both normal subjects and in patients with high blood pressure [61]. Therefore, an increased volume after renal 12.7 Potential components of the sodium retention would contribute to a state of arterial metabolic syndrome in the volume-dependent hypertension. elderly Finally, another possible explanation between IR and hypertension may also be linked to a 12.7.1 Oxidative stress derangement of cations. The ATPase Na/K+ pump is insulin-sensitive, and thus during IR a significant Recent research has focused on the progressive increase in intracellular sodium (Na+) accompanies a changes that occur in DNA structure in the metabolic consequent increase in intracellular calcium (Ca2+). syndrome, and the potential consequences of such During the metabolic syndrome an alteration in mutations. It has been suggested that excess and the haemostatic system occurs which results in a unopposed oxidative stress is a major cause of the potentially ‘pro-thrombotic state’ with severe vascular increased mitochondrial DNA (mtDNA) mutations complications. Three main components of the haemo- that occur with aging and metabolic syndrome. static system are activated during an array of vascular Oxidative stress is characterized by an uncontrolled tissue injury, including blood platelets, endothelial production of free radicals that are derived from cells and plasma coagulation factors. During an IR oxygen and produced by splitting covalent bond state, such reliability becomes less effective, which into atoms or molecules with an unpaired electron; shows that insulin is intrinsically involved in the this results in the formation of highly reactive correct operation of such a system. Therefore, any oxygen species (ROS). Under normal physiological alteration in the insulin signalling pathway has been conditions, the intra-mitochondrial environment is shown to result in an amplified platelet activation and characterized by a substantial equilibrium between the increased coagulation cascade activity, accompanied production of ROS and the activity of anti-oxidant by a parallel reduction in fibrinolysis. Insulin receptors mechanisms, such as glutathione peroxidase (GSH-Px) are normally present on the surface of blood platelets, and superoxide dismutase (SOD). However, when and when insulin binds to such receptors the insulin the endogenous production of ROS substantially signalling pathway is activated. Various studies have increases with a parallel decrease in anti-oxidant shown that insulin is capable of reducing the platelet agents, tissue damage will occur as result of such response to adenosine diphosphate (ADP), thrombin, oxidative stress. The results of various studies have adrenaline (epinephrine) and platelet-activating factor, suggested that the degree of unopposed oxidative as well as angiotensin II [62, 63]. Furthermore, stress is also predictive of mortality. In particular, activation of the insulin signalling pathway also the production of free radicals in the heart, kidney results in lower concentrations of calcium in the and liver is inversely proportional to the maximum platelets. Therefore, a normal insulin action has an lifespan [66], while the rate of mitochondrial oxygen inhibitory/regulatory action on platelet aggregation. In radical generation is negatively associated with animal an IR state, such equilibrium is lost and a persistent longevity. In animal models, calorie restriction also pro-thrombic condition occurs with an increased risk decreases mitochondrial oxygen radical production for vascular obstruction. Increased levels of fibrinogen and oxidative damage to mtDNA and decreasing the have been associated with IR. Fibrinogen is not rate of aging; some epidemiological studies have also only of fundamental importance in thrombin activity suggested that dietary anti-oxidants might have a regulation, but also has a predictive value for future significant impact on age-related disease states [67, cardiovascular events [64, 65]. In this context, it is 68]. The benefits of supplemental anti-oxidants remain important to underline that pro-inflammatory cytokine unproven in clinical trials. 172 CH 12 METABOLIC RISK FACTORS, OBESITY AND CARDIOMETABOLIC SYNDROME Oxidative stress also adversely impacts other vul- persons with two or more metabolic syndrome phe- nerable targets, including the lipid and protein compo- notypes had significantly higher homocysteine levels nents of membranes. Free radicals then allow for lipid compared to those with one or no metabolic syndrome oxidation and a consequent reduction in transmem- phenotypes [70]. Interestingly, elevated homocysteine brane transportation. This mechanism of uncontrolled levels increase the risk for CVD in type 2 diabetes oxidative stress (which is already active in older per- patients to a greater extent than among non-diabetic sons) becomes extremely active during the metabolic subjects. In a large community-based population of syndrome, and the link between IR and endothelial non-diabetic individuals, a modest association was dysfunction may explain, in part, the increased risk found between IR and elevated homocysteine levels for CVD associated with metabolic syndrome. The [71]. These authors found that the co-occurrence of mechanisms by which IR leads to endothelial dys- specific features of metabolic syndrome – especially function are certainly multiple and complex. All major hypertension and central obesity – was associated with abnormalities of the metabolic syndrome, such as hy- more marked elevations in homocysteine levels. Ho- perglycaemia, hypertension, dyslipidaemia and altered mocysteine as a contributory factor for vascular dam- coagulation, are also directly linked to endothelial dys- age remains controversial; however, the presence of function. However, with regard to the link between high plasma levels of homocysteine may play an IR and oxidative stress, it is important to underline important role and are capable of promoting oxida- recent observations. As noted above, insulin has a di- tive damage to the endothelium of vascular cells rect vasodilatory effect which is mediated through the through an auto-oxidation, the formation of homocys- stimulation of NO production in endothelial cells [69]. teine mixed disulfides and ROS formation [72]. In par- During an IR state, the ability to stimulate NO be- ticular, the oxidation of two homocysteine molecules comes limited, while at the same time an increase in results in the formation of oxidized disulfide, two ROS, such as superoxide, occurs. Pro-inflammatory cy- protons and two electrons, while promoting the for- tokines (e.g. TNF-α and IL-6) function synergistically mation of ROS. The activation of ROS is only one with ROS towards creating endothelial disarray and of the many damaging effects of homocysteine, as it thus, an increased risk for the development of vascular may act either alone or in concert with other multi- disease. Therefore, it might be speculated that individ- ple injurious stimuli to damage endothelial cell func- uals who exhibit the metabolic syndrome may have tion. The combined effect of hypercoagulabilty and an abnormality in their NO production by endothelial IR strengthen the processes of oxidation, glycation or cells and a simultaneous and constant stimulation of homocysteinylation of LDLs necessary for the trans- pro-atherogenic changes in the vasculature in response formation in atherogenic particles. There are multiple to IR. metabolic toxicities associated with metabolic syndrome (and in particular IR) which are also associated with the production of ROS, thus creating multi- 12.7.2 Homocysteine ple injurious stimuli and a higher risk for accelerated atherosclerosis. Homocysteine, a sulphur-containing amino acid which is formed during the metabolism of methionine, has emerged as a novel independent biomarker for the 12.7.3 Cognitive functioning development of atherosclerotic disease in coronary, The aging process is associated with a significant de- cerebral and peripheral vascular beds. In fact, there cline in cognitive functioning. Indeed, many studies is an increased risk for CVD in those with elevated have attempted to discover the mechanisms involved fasting plasma total homocysteine levels. It has been with cognitive decline in vulnerable older persons. shown that those individuals with hyperinsulinaemia Many of the same risk factors associated with CVD (which is considered a marker of IR) had signif- have also been recognized with a higher risk for de- icantly higher homocysteine levels than those with veloping compromised cognitive functioning. The risk normal insulin levels. The plasma levels of insulin of cognitive impairment, especially in older persons seem to influence homocysteine metabolism, possi- with metabolic syndrome, needs to be recognized by bly through the effects on glomerular filtration or by physicians, due to the fact that its association with a influencing the activity of key enzymes in homocys- lower functional status in older persons opens new teine metabolism. These authors also confirmed that public health concerns. The metabolic syndrome and 12.8 THERAPEUTIC PERSPECTIVES 173 its individual components have also been associated intracellular glucose metabolism. Thus, it is plausible with an increased risk of developing cognitive impair- that age-related IR may be a determinant of reduced ment and decline over four years in high-functioning muscle functioning, as seen clinically by a lower mus- older persons, even after adjusting for comorbidities cle strength. One cross-sectional study demonstrated a [73]. In particular, such decline was steeper in those significant association between IR and muscle strength who had high serum levels of inflammatory markers. in older non-diabetic persons, independent of multiple Age-related IR has been shown to be independently confounders [81]. In particular, insulin may also be an associated with reduced cognitive functioning in older, important determinant of muscle function as glucose non-diabetic persons [74]. Indeed, it is now well doc- uptake is necessary for adequate muscle contraction. umented that insulin is a fundamental neuromodulator, Another important role played by insulin is its ability contributing to neurobiological processes in particular, to repress whole-body proteolysis, thus shifting total energy homeostasis and cognition. Interestingly, in- body metabolism towards an anabolic state. Therefore, sulin and insulin receptors are selectively expressed in it is plausible that an age-related IR may be a deter- the brain [75]. Initially, the majority of glucose trans- minant of poor muscle strength in older persons with porters in the brain were considered insulin-insensitive, metabolic syndrome. Furthermore, a reduction of in- but it has been shown recently that a significant ele- sulin peripheral activity may reduce the muscle tissue ment of brain glucose uptake is insulin-sensitive and anabolic rate, leading to a relative catabolic state and, essential for correct cognitive functioning [76]. Thus, in turn, contributing to the impairment in muscle func- an age-related reduction in glucose uptake due to al- tioning. Indeed, insulin is known to play a pivotal role tered insulin-signalling can lead to a deficiency of en- in muscle functioning by increasing glucose uptake and ergetic substrate that cannot be compensated by other promoting intracellular glucose metabolism. metabolic pathways [77]. Interestingly, a significant increase in peripheral IR develops as individuals age, thus raising the possibility that a reduced efficiency 12.8 Therapeutic perspectives of the metabolic pathway responsible for energy production is one of the mechanisms of cognitive decline The metabolic risk factors of the metabolic syndrome in older persons [78, 79]. Collectively, these findings include atherogenic dyslipidaemia, hypertension, IR, suggest that: (i) insulin contributes to normal cognitive a prothrombotic state and the pro-inflammatory state. functioning; and (ii) insulin abnormalities exacerbate Due to the fact that IR is the key player in the metabolic cognitive impairment, such as those associated with syndrome, drug interventions should be used either di- Alzheimer’s disease. It is noteworthy that endothe- rectly to improve insulin sensitivity, or indirectly to lial damage and vascular disease, when combined with improve the metabolic changes associated with IR. IR, are also responsible for an age-related decline in All risk factors linked to IR must be treated in or- cognitive function, even in the absence of dementia. der to reduce the severity of metabolic syndrome (see Furthermore, any factors capable of lowering or in- Table 12.4). Both, the American Diabetes Associa- creasing the risk of endothelial damage and/or affecting tion (ADA) [83] and the National Cholesterol Edu- insulin action may have a role on cognitive function cation Program (NCEP) have adopted guidelines for and cerebrovascular diseases. complications related to metabolic syndrome [2]. The cornerstones of metabolic syndrome treatment are the 12.7.4 Muscle functioning management of body weight, and ensuring that appropriate physical exercise is performed. In fact, education In older persons, poor muscle strength and poor phys- and training should be considered fundamental, due to ical performance often coexist. the fact that environmental influences such as incorrect Midlife handgrip muscle strength has been recog- nutrition and physical inactivity are considered to be nized as an important factor that predicts old age func- root causes of the metabolic syndrome. Interventions tional ability [80]. Observational studies have consis- on the metabolic syndrome with physical activity have tently shown that chronic conditions such as CHD, dia- shown that regular and sustained physical activity will betes and pulmonary obstructive disease are associated improve all of the risk factors [84, 85]. With regards to with lower muscle strength [81, 82]. It is also widely weight management, there is a general consensus that known that insulin plays a pivotal role for muscle con- persons with metabolic syndrome should reduce their traction by increasing glucose uptake and promoting consumption of simple sugars and increase their intake 174 CH 12 METABOLIC RISK FACTORS, OBESITY AND CARDIOMETABOLIC SYNDROME

Table 12.4 Theraneutic approach to the various components of the metabolic syndrome. Therapeutic approach Abdominal obesity Dyslipidaemia Hypertension Impaired glucose tolerance Diet Reduces calories Reduces unsatured fat Reduces sodium intake Reduces glucose intake – Daily activity/exercise – Daily activity/exercise – Daily activity/exercise – Daily activity/exercise – Fruits –Omega-3s – Fruits –Fibre – Vegetables –MUFA – Vegetables – Fruits – Whole grains –Satfat – Whole grains – Vegetables – Trans fat – Whole grains Medication – Rimonabant – Statin – Angiotensin-Converting – Metformin – Fibrates Enzyme Inhibitors – Acarbose – Angiotensin receptor – Incretin? blocker – DPP-IV inhibitor? – Thiazolidinediones?

of fruits, vegetables and whole grains [86]. Although, and by stimulating the insulin-signalling pathway. Var- the recommended intakes of carbohydrate and unsatu- ious studies have also confirmed that PPAR-γ agonists rated fats remain controversial [87], it has been shown are capable of improving the altered lipid metabolism that low-fat diets promote weight reduction, while a associated with IR. PPAR-γ is highly expressed in adi- higher monosaturated fat intake reduces postprandial pose tissue, where it triggers adipocyte differentiation glycaemia, reduces plasma triglyceride levels and also as well as inducing genes that are critical for adipoge- raises HDL concentrations [88]. Although modifica- nesis. Adipose tissue serves as a major site of oxidized tions to diet and exercise need to be adapted in the first LDL (oxLDL) detoxification, causing it to be removed approach for ameliorating the metabolic syndrome, the from the bloodstream and potentially inhibiting the for- use of pharmaceuticals is almost always necessary. mation of atherosclerotic lesions. Increasing clinical Thiazolidinediones (TZD or glitazones) represent a evidence has also shown that rosiglitazone treatment new class of oral antidiabetic drug which exert their significantly improves factors associated with CVD, insulin-sensitizing action by stimulating the nuclear including endothelial activity, inflammatory processes transcription factor peroxisome proliferator-activated and dyslipidaemia [90, 91]. Only future clinical tri- receptor gamma (PPAR-γ). At present, pioglitazone als conducted in older persons in an IR state will and rosiglitazone are available for clinical use. The be able to clarify whether the protective factors from different activation levels of PPAR-γ and of their co- the use of TZDs are substantial. TZDs are also ca- factors determine the binding of PPAR-γ to distinct pable of reducing oxidized LDLs in both lean and target genes, which in turn regulates their transcrip- obese diabetic animals [92]. In particular, TZDs have tional activity. It is important to recall that PPARs are been shown to up-regulate the oxidized LDL receptor members of the superfamily of nuclear hormone re- 1 (OLR1) in adipocytes by facilitating the exchange ceptors; these are transcription factors which transmit of coactivators for corepressors on the OLR1 gene signals that originate from lipid-soluble factors to the in cultured mouse adipocytes. TZDs markedly stim- genome. Nuclear receptors bind to DNA at specific ulated the uptake of oxLDL into adipocytes, which sites or response elements, that in turn can activate or required OLR1. Increased OLR1 expression, result- repress the expression of a target gene. Three different ing from TZD treatment, significantly increased the PPAR genes (α, γ and δ) have been identified [89] adipocyte cholesterol content and also enhanced FA which play distinct expression patterns, and this sug- uptake. While the physiological role of adipose tis- gests that they have important functional differences. sue in cholesterol and oxLDL metabolism remains TZDs dramatically upgrade insulin sensitivity through unknown, the induction of OLRI may be a potential PPAR-γ , and are also capable of lowering blood glu- means through which PPAR-γ ligands may regulate cose levels through more active glucose transporters lipid metabolism and insulin sensitivity in adipocytes 12.8 THERAPEUTIC PERSPECTIVES 175 [93]. A degree of caution should be applied when treat- decreases food intake and body weight while ing older persons with PPAR-γ agonists, due to the fact increasing adiponectin and insulin sensitivity. Drug that these agents may cause water retention and an al- interventions with rimonabant (e.g. the RIO-Europe tered hepatic detoxifying activity; this in turn would study) addressed the effects of such therapy on weight exacerbate a potential condition of cardiac failure and loss, altered glucose metabolism and dyslipidaemia an inadequate removal of toxic compounds due to an [97]. It was found that, at the one-year follow-up inappropriate liver metabolism. PPAR-α agonists or fi- examination, those patients receiving 20 mg per brates are used to treat dyslipidaemia, particularly in day rimonabant showed significant reductions in the case of high triglycerides and low HDL-cholesterol. both body weight and waist circumference, as well Fibrates (gemf`ıbrozil, bezaf`ıbrate, fenofibrate) rep- as significant improvements in lipid and glucose resent another class of lipid-lowering drugs known parameters. A significant increase was noted in HDL generally to be effective for reducing elevated plasma levels and a decrease in triglycerides, as well as triglyceride and cholesterol levels. Although very few a significant reduction in IR, as monitored using controlled clinical trials have been conducted compar- the homeostasis model assessment index (HOMA). ing fibrates with statins (especially in older persons), the use of fibrates has been shown to include protective The risk for cardiovascular events are significantly effects towards further vascular events in survivors of increased during an IR state, due to the latter being myocardial infarction [94]. The more pronounced ef- ameliorated by an enhanced lipid metabolism; hence, fect of fibrates is a decrease in plasma triglyceride-rich lipid-lowering agents such as statins may have lipoproteins after linking to PPAR-α. The latter is es- an important role. It is widely known that statins pecially expressed in kidney, heart and muscle tis- (3-hydroxy-3-methylglutaryl-coenzyme A reductase sue (all of which metabolize large amounts of fatty inhibitors) lower the risk of CVD by reducing the acids). Fibrates are ligands for PPAR-α [95]; thus, their production of all apo B-containing lipoproteins and main mechanisms of action are linked to the activa- VLDL. Statins are effective, in both the primary and tion of key genes involved in lipid metabolism. The secondary prevention of CHD, in middle-aged and hypotriglyceridaemic action of fibrates involves a com- older (<65 years) men and women, in both diabetics bination effect of HL and apoC-III expression. HL is and non-diabetics with CHD [98, 99]. Statins used in induced at the transcriptional level mediated by PPAR, the secondary prevention of CHD also significantly while apoC-III is repressed. As a consequence a re- reduce not only the risk of stroke but also the duced secretion of VLDL particles occurs, together frequency of daily attacks of myocardial ischaemia. with an enhanced catabolism of triglyceride-rich par- Statins have been shown to reduce LDL levels by ticles. Fibrates are also known to increase the hepatic 20–50%, triglycerides by 10–40%, and to increase uptake of FFAs, to increase the removal of LDL par- HDL levels by 5–12%. In fact, the Adult Treatment ticles, and to stimulate the production of HDL and its Panel III (ATP III) of the National Cholesterol major constituents, apoA-I and apoA-II. Hypertriglyc- Education Program have issued evidence-based eridaemia, in association with an IR state, is considered guidelines of major clinical trials with statin therapy. the main target for the use of fibrates in the metabolic According to the ATP III algorithm, persons may be syndrome. In fact, in those persons undergoing therapy allocated to one of three risk categories: with fenofibrate, a 50% decrease in triglycerides levels was noted, together with a simultaneous 10–30% increase in basal HDL levels [96]. Fenofibrate was also 1. Established CVD and CVD risk equivalents. found to be more effective in reducing the plasma con- 2. Multiple (two or more risk factors). centrations of oxLDLs, while in individuals at high risk of CVD the impact of accelerating chylomicron and 3. Zero to one risk factor. VLDL catabolism underlined the ability of fibrates to act on postprandial lipid metabolism [94]. Only large Here, the CVD risks include non-coronary forms of clinical studies will confirm the efficacy of fibrates atherosclerotic disease, diabetes and multiple (two or in reducing coronary events and mortality in older, more) CVD risk factors with a 10-year risk for CVD high-risk individuals. exceeding 20%. Therefore, all persons with CVD or A very recently introduced drug, rimonabant – a CVD risk equivalents may be considered at high risk selective cannabinoid-1 (CB1) receptor blocker – of CVD. 176 CH 12 METABOLIC RISK FACTORS, OBESITY AND CARDIOMETABOLIC SYNDROME 12.8.1 Therapy in high-risk CVD patients loss. However, if the hypertension persists despite such changes, then anti-hypertensive drug treatment is usu- The goal for LDL-lowering therapy is to achieve ally required. It is widely known that the benefits dean LDL level <100 mg dl−1 in high-risk patients; rived from a reduction in arterial hypertension lower therefore, in persons with an LDL level <100 mg dl−1 the risk for CVD [104]. It also has been suggested no further LDL-lowering therapy is recommended, that the use of angiotensin-converting enzyme (ACE) whereas in high-risk patients with an LDL level inhibitors or angiotensin receptor blockers represent >100 mg dl−1 a diet-based therapy should be initi- a better first-line therapy for metabolic syndrome pa- ated. When the baseline LDL is >130 mg dl−1,the tients [105]. The atherothrombotic state characterized LDL-lowering drug should be commenced simultane- by high concentrations of fibrinogen, PAI-1 and in- ously with diet-based therapy. Then, if high triglyc- creased platelet aggregation may be treated when nec- eride or low HDL levels persist, consideration may essary with low-dose aspirin or other anti-platelet drugs given to the use of a fibrate. [106]. Indeed, such drugs are universally recommended in patients with established CVD, although their effi- 12.8.2 Therapy in moderate-risk CVD cacy in older persons with metabolic syndrome in the patients absence of CVD has still to be verified in clinical trials. Moderate-risk persons have two or more risk However, the use of aspirin has been considered a pro- factors and 10-year risk of between 10% and 20%. phylactic option when the risk for CVD is high [107]. The recommendation in this case is to obtain an LDL level <130 mg dl−1. When the LDL level is 100–129 mg dl−1, either at baseline or on lifestyle 12.9 Conclusions therapy, the initiation of an LDL-lowering drug to achieve LDL <100 mg dl−1 is therapeutic option. Today, the proportion of aged subjects among the When LDL-lowering drug treatment is initiated in populations of industrialized countries continues to high-risk or moderate-risk patients, it is advised that increase dramatically, thus predisposing such persons the goal of such therapy is to achieve at least a to development of the metabolic syndrome. If it is 30–40% reduction in LDL levels. accepted that each single component of the metabolic The question remains, however, as to how exactly syndrome is significantly linked to IR, then it will be statins improve IR. It has been reported in several vital for physicians to focus on correcting IR in order studies that statins also have a beneficial impact on to improve each of these metabolic components. the chemical properties of lipoproteins by reducing not only oxLDLs but also small and dense VLDLs, all of which are commonly observed during an IR state [100]. In particular, small doses of atorvastatin (10 mg) have References been shown to reduce postprandial concentrations of VLDLs, IDLs and apo B in persons with normal lipid 1. Reaven GM. Banting lecture 1988. Role of insulin profiles [101]. Furthermore, in patients with altered resistance in human disease. Diabetes 1988; 37: lipid profiles and altered glucose metabolism (reduced 1595–607. 2. Executive Summary of the Third Report of the Na- glucose tolerance or type 2 diabetes), treatment with tional Cholesterol Education Program (NCEP) Expert atorvastatin led to reductions in both IR (confirmed Panel on Detection, Evaluation, and Treatment of using the HOMA index [102]) and small, dense athero- High Blood Cholesterol in Adults (Adult Treatment genic LDLs [103]. Panel III). JAMA 2001; 285: 2486–97. Simvastatin, rosuvastatin and fluvastatin have also 3. Isomaa B, Almgren P, Tuomi T, Forsen B, Lahti K, been shown to reduce oxLDLs, thus generalizing the Nissen M, Taskinen MR and Groop L. Cardiovascular so-called ‘class effect’ of statins. It has been hypothe- morbidity and mortality associated with the metabolic sized that a reduction in HL activity during such ther- syndrome. Diabetes Care 2001; 24: 683–9. apy may be at the basis of such a phenomenon. With 4. WHO. (1999) Definition of metabolic syndrome in regards to arterial hypertension and IR, mild elevations definition, diagnosis and classification of diabetes of blood pressure can often be controlled by lifestyle and its complications. Report of a WHO consulta- changes, including a reduced sodium intake and weight tion. Part 1: Diagnosis and classification of diabetes 12.9 CONCLUSIONS 177

mellitus. WHO/NCD/NCS/99.2. World Health Orga- 18. Gumbiner B, Thorburn AW, Ditzier TM, Bulacan F nization, Department of Noncommunicable Disease and Henry RR. Role of impaired intracellular glu- Surveillance, Geneva. cose metabolism in the insulin resistance of aging. 5. Bloomgarden ZT. American Association of Clinical Metabolism 1992: 41 (10): 1115–21. Endocrinologists (AACE) consensus conference on 19. Ferrannini E, Vichi S, Beck-Nielsen H, Laakso M, the insulin resistance syndrome: 25–26 August 2002, Paolisso G and Smith U. Insulin action and age. Washington, DC. Diabetes Care 2003; 26: 1297–303. European Group for the Study of Insulin Resistance 6. Einhorn D, Reaven GM, Cobin RH, Ford E, Ganda (EGIR). Diabetes 1996: 45 (7): 947–53. OP, Handelsman Y, Hellman R, Jellinger PS, Kendall 20. Barbieri M, Rizzo MR, Manzella D and Paolisso G. D, Krauss RM, Neufeld ND, Petak SM, Rodbard Age-related insulin resistance: is it an obligatory find- HW, Seibel JA, Smith DA and Wilson PW. American ing? The lesson from healthy centenarians. Diabetes College of Endocrinology position statement on the Metab Res Rev. 2001; 17 (1): 19–26. insulin resistance syndrome. Endocr Pract. 2003; 9: 21. Banerji MA, Chaiken RL, Gordon D, Kral JG 237–52. and Lebovitz HE. Does intra-abdominal adipose 7. Yach D, Stuckler D and Brownell KD. Epidemiologic tissue in black men determine whether NIDDM is and economic consequences of the global epidemics insulin-resistant or insulin-sensitive? Diabetes 1995; of obesity and diabetes. Nat Med. 2006; 12: 62–6. 44 (2): 141–6. 8. Bonadonna RC, Cucinotta D, Fedele D, Riccardi G, 22. Fujimoto WY, Abbate SL, Kahn SE, Hokanson JE Tiengo A and the Metascreen Writing Committee. and Brunzell JD. The visceral adiposity syndrome in The metabolic syndrome is a risk indicator of mi- Japanese-American men. Obes Res 1994; 2: 364–71. crovascular and macrovascular complications in dia- 23. Laws A, Hoen HM, Selby JV, Saad MF, Haffner SM betes: results from Metascreen, a multicenter diabetes and Howard BV. Differences in insulin suppression of clinic-based survey. Diabetes Care 2006; 29: 2701–7. free fatty acid levels by gender and glucose tolerance 9. Grundy SM. Metabolic syndrome: a multiplex cardio- status. Relation to plasma triglyceride and apolipopro- vascular risk factor. J Clin Endocrinol Metab. 2007; tein B concentrations. Insulin Resistance Atheroscle- 92: 399–404. rosis Study (IRAS) Investigators. Arterioscler Thromb 10. Pi-Sunyer FX. The relation of adipose tissue to car- Vasc Biol. 1997; 17 (1): 64–71. diometabolic risk. Clin Cornerstone 2006; 8 (Suppl. 24. Jeppesen J, Hein HO, Suadicani P and Gyntelberg 4): S14–23. F. Triglyceride concentration and ischemic heart dis- 11. Katagiri H, Yamada T and Oka Y. Adiposity and ease: an eight-year follow-up in the Copenhagen Male cardiovascular disorders: disturbance of the regulatory Study. Circulation 1998; 97 (11): 1029–36. system consisting of humoral and neuronal signals. 25. Zambon A, Austin MA, Brown BG, Hokanson JE Circ Res. 2007; 101: 27–39. and Brunzell JD. Effect of hepatic lipase on LDL in 12. Lee M and Aronne LJ. Weight management for type 2 normal men and those with coronary artery disease. diabetes mellitus: global cardiovascular risk reduction. Arterioscler Thromb. 1993; 13 (2): 147–53. Am J Cardiol. 2007; 99 (4A): 68B–79B. 26. Morton RE and Ziiversmit DB. Purification and char- 13. Bloomgarden ZT. Insulin resistance concepts. Dia- acterization of lipid transfer protein(s) from human betes Care 2007; 30: 1320–6. lipoprotein-deficient plasma. J Lipid Res. 1982; 23 14. Rader DJ. Effect of insulin resistance, dyslipidemia, (7): 1058–67. and intra-abdominal adiposity on the development of 27. Krauss RM. Dense low density lipoproteins and coro- cardiovascular disease and diabetes mellitus. Am J nary artery disease. Am J Cardiol. 1995; 75 (6): Med. 2007; 120 (3 Suppl. 1): S12–18. 53B–57B. 15. Sarti C and Gallagher J. The metabolic syndrome: 28. Rashid S, Uffelman KD and Lewis GF. The mech- prevalence, CHD risk, and treatment. J Diabetes Com- anism of HDL lowering in hypertriglyceridemic, plications 2006; 20: 121–32. insulin-resistant states. J Diabetes Complications 16. Fink RI, Kolterman OG, Griffin J and Olefsky JM. 2002; 16 (1): 24–8. Mechanisms of insulin resistance in aging. J din 29. Clay MA, Newnham HH and Barter PJ. Hepatic Invest. 1983: 71 (6): 1523–35. lipase promotes a loss of apolipoprotein A-I from 17. Rowe JW, Minaker KL, Pallotta JA and Flier JS. triglyceride-enriched human high density lipoproteins Characterization of the insulin resistance of aging. J during incubation in vitro. Arterioscler Thromb. 1991; Clin Invest. 1983: 71 (6): 1581–7. 1 (2): 415–22. 178 CH 12 METABOLIC RISK FACTORS, OBESITY AND CARDIOMETABOLIC SYNDROME

30. Austin MA, King MC, Vranizan KM and Krauss RM. lipoproteins: evidence for independent roles of age Atherogenic lipoprotein phenotype. A proposed ge- and sex. Diabetologia 2003; 46 (4): 459–69. netic marker for coronary heart disease risk. Circula- 44. Salmenniemi U, Ruotsalainen E, Pihiajamaki J, tion 1990; 82 (2): 495–506. Vauhkonen I, Kainulainen S, Punnonen K, Vanninen 31. Pierson RN Jr. Body composition in aging: a bio- E and Laakso M. Multiple abnormalities in glucose logical perspective. Curr Opin Clin Nutr Metab Care and energy metabolism and coordinated changes 2003; 6 (1): 15–20. in levels of adiponectin, cytokines, and adhesion 32. Kannel WB, Cupples LA, Ramaswami R, Stokes J, molecules in subjects with metabolic syndrome III, Kreger BE and Higgins M. Regional obesity and Circulation 2004; 110 (25): 3842–8. risk of cardiovascular disease; the Framingham Study. 45.OuchiN,KiharaS,AritaY,MaedaK,KuriyamaH. Epidemiology 1991: 44 (2): 183–90. Okamoto Y, Hotta K, Nishida M, Takahashi M, Naka- 33. Harris TB, Launer LJ, Madans J and Feldman JJ. mura T, Yamashita S, Funahashi T and Matsuzawa Y. Cohort study of effect of being overweight and change Novel modulator for endothelial adhesion molecules: in weight on risk of coronary heart disease in old age. adipocyte-derived plasma protein adiponectin. Circu- Br Med J. 1997; 314 (7097): 1791–4. lation 1999; 100: 2473–6. 34. Elias MF, Elias PK, Sullivan LM, Wolf PA and 46. Okamoto Y, Arita Y and Nishida M. An D’Agostino RB. Lower cognitive function in the pres- adipocyte-derived plasma protein, adiponectin, ence of obesity and hypertension: the Framingham adheres to injured vascular walls. Horm Metab Res Heart Study. Int J Obes Relat Metab Disord. 2003; 2000; 32: 47–50. 27 (2): 260–8. 47. Pischon T, Girman CJ, Hotamisligil GS, Rifai N, Hu 35. Chumlea WC, Garry PJ, Hunt WC and Rhyne RL. FB and Rimm EB. Plasma adiponectin levels and risk Distributions of serial changes in stature and weight of myocardial infarction in men. JAMA 2004: 291: in a healthy elderly population. Hum Biol. 1988; 60 1730–7. (6): 917–31. 48. Maahs DM, Ogden LG, Kinney GL, Wadwa P, 36. Ahima RS and Flier JS. Adipose tissue as an en- Snell-Bergeon JK, Dabelea D, Hokanson JE, Ehriich docrine organ. Trends Endocrinol Metab 2000; 11: J, Eckel RH and Rewers M. Low plasma adiponectin 327–32. levels predict progression of coronary artery calcifi- 37. Rajala MW and Scherer PE. Minireview: The cation. Circulation 2005; 11: 747–53. adipocytes at the crossroads of energy homeostasis, 49. Paolisso G, Rizzo MR, Mazziotti G, Tagliamonte MR, inflammation, and atherosclerosis. Endocrinology Gambardella A. Rotondi M, Carella C, Giugliano 2003; 144 (9): 3765–73. D, Varricchio M and D’Onofrio F. Advancing age 38. Lyon CJ and Law WA. Minireview: Adiposity, in- and insulin resistance: role of plasma tumor necrosis flammation, and atherogenesis. Endocrinology 2003; factor-alpha. Am J Physiol 1998; 275: E294–9. 144 (6): 2195–200. 50. Abbatecola AM, Ferrucci L, Grella R, Bandinelli 39. Fasshauer M and Paschke R. Regulation of adipocy- S, Bonafe M, Barbieri M, Corsi AM, Lauretani tokines and insulin resistance. Diabetologia 2003; 46 F, Franceschi C and Paolisso G. Diverse effect (12): 1594–603. of inflammatory markers on insulin resistance and 40. Trayhurn P and Wood IS. Adipokines: inflammation insulin-resistance syndrome in the elderly. J Am Geri- and the pleiotropic role of white adipose tissue. Br J atr Soc 2004; 52: 399–404. Nutr. 2004; 92 (3): 347–55. 51. Franceschi C, Bonafe M, Valensin S, Olivieri F, 41. Yudkin JS, Stehouwer CD, Emeis JJ and Coppack De Luca M, Ottaviani E and De Benedictis G. SW. C-reactive protein in healthy subjects: associa- Inflamm-aging. An evolutionary perspective on im- tions with obesity, insulin resistance, and endothelial munosenescence. Ann N Y Acad Sci. 2000; 908: dysfunction. A potential role for cytokines originating 244–54. from adipose tissue? Arterioscler Thromb Vasc Biol 52. Hak AE, Pois HAP, Stehouwer CDA, et al.Markers 1999; 19: 972–8. of inflammation and cellular adhesion molecules in 42. Chandran M, Phillips SA, Ciaraldi T and Henry RR. relation to insulin resistance in nondiabetic elderly: Adiponectin: more than just another fat cell hormone? The Rotterdam study. J Clin Endocrinol Metab 2001; Diabetes Care 2003; 26 (8): 2442–50. 86: 4398–405. 43. Cnop M, Havei PJ, Utzschneider KM, Carr DB, Sinha 53. Welborn TA, Breckenridge A, Rubinstein AH, Dollery MK, Boyko EJ, Retzlaff BM, Knopp RH, Brunzell CT and Fraser TR. Serum-insulin in essential hy- JD and Kahn SE. Relationship of adiponectin to pertension and in peripheral vascular disease. Lancet body fat distribution, insulin sensitivity and plasma 1966; 18: 1336–7. 12.9 CONCLUSIONS 179

54. Modan M, Halkin H, Almog S, Lusky A, Eshkol A, 65. Anand SS, Yi Q, Gerstein H, Lonn E, Jacobs R, Vuk- Shef`ı M, Shitrit A and Fuchs Z. Hyperinsulinemia. san V, Teo K, Davis B, Montague P and Yusuf S. Re- A link between hypertension obesity and glucose lationship of metabolic syndrome and fibrinolytic dys- intolerance. J Clin Invest. 1985; 75: 809–17. function to cardiovascular disease. Circulation 2003; 55. Zavaroni I, Bonini L, Gasparini P, Barilli AL, Zuc- 108: 420–5. carelli A, Dall’Aglio E, Del Signore R and Reaven 66. Sohai RS, Svensson I, Sohai BH and Brunk UT. Su- GM. Hyperinsulinemia in a normal population as a peroxide anion radical production in different animal species. Mech Ageing Dev. 1989; 49: 129–35. predictor of non-insulin-dependent diabetes mellitus, 67. Gilgun-Sherki Y, Melamed E and Offen D. Antioxi- hypertension, and coronary heart disease: the Barilla dant treatment in Alzheimer’s disease: current state. J factory revisited. Metabolism 1999; 48: 989–94. Mol Neurosci. 2003; 21: 1–12. 56. Ferrannini E, Buzzigoli G, Bonadonna R, Giorico 68. Paolini M, Sapone A, Canistro D, Chicco P and MA, Oleggini M, Graziadei L, Pedrinelli R, Brandi Valgimigli L. Antioxidant vitamins for prevention of L and Bevilacqua S. Insulin resistance in essential cardiovascular disease. Lancet 2003; 362: 920–4. hypertension. N Engl J Med 1987; 317: 350–7. 69. Kuboki K, Jiang ZY, Takahara N, Ha SW, Igarashi 57. Steinberg HO, Brechtel G, Johnson A, Fineberg N M, Yamauchi T, Feener EP, Herbert TP, Rhodes CJ and Baron AD. Insulin-mediated skeletal muscle va- and King GL. Regulation of endothelial constitutive sodilation is nitric oxide dependent. A novel action of nitric oxide synthase gene expression in endothelial insulin to increase nitric oxide release. J Clin Invest. cells and in vivo: a specific vascular action of insulin. 1994; 94: 1172–9. Circulation 2000; 101: 676–81. 58. Steinberg HO, Chaker H, Leaming R, Johnson A, 70. Meigs JB, Jacques PF, Seihub J, Singer DE, Nathan Brechtel G and Baron AD. Obesity/insulin resistance DM, Rifai N, D’Agostino RB, Sr and Wilson PW. is associated with endothelial dysfunction. Implica- Fasting plasma homocysteine levels in the insulin re- tions for the syndrome of insulin resistance. J Clin sistance syndrome: the Framingham Offspring Study. Invest. 1996; 97: 2601–10. Diabetes Care 2001; 24: 1403–10. 71. Hoogeveen EK, Kostense PJ, Beks PJ, Mackaay 59. Maaten JC, Voorburg A, de Vries PM, ter Wee PM, AJ, Jakobs C, Bouter LM, Heine RJ and Stehouwer Donker AJ and Gans RO. Relationship between in- CD. Hyperhomocysteinemia is associated with an sulin’s haemodynamic effects and insulin-mediated increased risk of cardiovascular disease, especially glucose uptake. Eur J Clin Invest. 1998; 28: 279–84. in non-insulin-dependent diabetes mellitus: a 60. Grassi G, Seravalle G, Cattaneo BM, Bolla GB, Lan- population-based study. Arterioscler Thromb Vasc franchi A, Colombo M, Giannattasio C, Brunani A, Biol. 1998; 18: 133–8. Cavagnini F and Mancia G. Sympathetic activation in 72. Jakubowski H, Zhang L, Bardeguez A and Aviv A. obese normotensive subjects. Hypertension 1995; 25: Homocysteine thiolactone and protein homocysteiny- 560–3. lation in human endothelial cells: implications for 61. DeFronzo RA, Cooke CR, Andres R, Falcona GR and atherosclerosis. Circ Res. 2000; 87: 45–51. Davis PJ. The effect of insulin on renal handling of 73. Yaffe K, Kanaya A, Lindquist K, Simonsick E, Harris sodium, potassium, calcium, and phosphate in man. J T, Shorr R, Tylavsky F and Newman A. The metabolic Clin Invest. 1975; 55: 845–55. syndrome, inflammation, and risk of cognitive decline. 62. Zavaroni I, Coruzzi P, Bonini L, Mossini GL, Musiari JAMA 2004; 292: 2237–42. L, Gasparini P, Fantuzzi M and Reaven GM. Associa- 74. Abbatecola AM, Paolisso G, Lamponi M, Bandinelli tion between salt sensitivity and insulin concentrations S, Lauretani F, Launer L and Ferrucci L Insulin in patients with hypertension. Am J Hypertens. 1995; resistance and executive dysfunction in older persons. J Am Geriatr Soc. 2004; 52: 1713–18. 8: 855–8. 75. Kyriaki G. Brain insulin: Regulation, mechanisms of 63. Trovati M and Anfossi G. Insulin, insulin resistance action and functions. Cellular and Molecular Neuro- and platelet function: similarities with insulin effects biology 2003; 23: 1–25. on cultured vascular smooth muscle cells. Diabetolo- 76. McEwen G and Reagan F. Glucose transporter ex- gia 1998; 41: 609–22. pression in the central nervous system: relationship to 64. Trovati M, Mularoni EM, Burzacca S, Ponziani MC, synaptic function. European Journal of Pharmacology Massucco P, Mattiello L, Piretto V, Cavalot F and An- 2004; 490: 13–24. fossi G. Impaired insulin-induced platelet antiaggre- 77. Messier C, Awad N and Gagnon M. The relationships gating effect in obesity and in obese NIDDM patients. between atherosclerosis, heart disease, type 2 diabetes Diabetes 1995; 44: 1318–22. and dementia. Neurol Res. 2004; 26: 567–72. 180 CH 12 METABOLIC RISK FACTORS, OBESITY AND CARDIOMETABOLIC SYNDROME

78. Craft S, Dagogo-Jack SE, Wiethop BV, Murphy C, prevention and management of type 2 diabetes: NevinsRT,FleischmanS,RiceV,NewcomerJW rationale and strategies. A statement of the American and Cryer PE. Effects of hyperglycemia on memory Diabetes Association, the North American Associ- and hormone levels in dementia of the Alzheimer ation for the Study of Obesity, and the American type: a longitudinal study. Behav Neurosci 1993; 107: Society for Clinical Nutrition. Am J Clin Nutr. 2004: 926–40. 80: 257–63. 79. Kalmijn S, Feskens EJM, Launer LJ, Stijnen T and 89. Schoonjans K, Staels B and Auwerx J. Role of the Kromhout D. Glucose intolerance, hyperinsulinemia peroxisome proliferator-activated receptor (PPAR) in and cognitive function in a general population of mediating the effects of fibrates and fatty acids on elderly men. Diabetologia 1995; 38: 1096–102. gene expression. J Lipid Res. 1996; 37: 907–25. 80. Rantanen T, Guralnik JM, Foley D, et al. Midlife hand 90. Gilling E, Suwattee P, DeSouza C, Asnani S and grip strength as a predictor of older age disability. Fonseca V. Effects of the thiazolidinediones on car- JAMA 1999: 10; 281: 558–60. diovascular risk factors. Am J Cardiovasc Drugs 2002; 81. Kalirnan DA, Piato CC and Tobin JD. The role of 2: 149–56. muscle loss in the age-related decline of grip strength: 91. Sidhu JS, Cowan D and Kaski JC. The effects cross-sectional and longitudinal perspectives. J Geron- of rosiglitazone, a peroxisome proliferator-activated tol. 1990: 45: M82–8. receptor-gamma agonist, on markers of endothelial 82. Abbatecola AM, Ferrucci L, Ceda G, Russo CR, cell activation. C-reactive protein, and fibrinogen lev- Lauretani F, Bandinelli S, Barbieri M, Valenti G and els in non-diabetic coronary artery disease patients. J Paolisso G. Insulin resistance and muscle strength in Am Coll Cardiol. 2003; 42: 1757–63. older persons. J Gerontol A Biol Sci Med Sci. 2005; 92. Buse JB, Tan MH, Prince MJ and Erickson PP. The 60: 1278–82. effects of oral anti-hyperglycaemic medications on 83. Franz MJ, Bantle JP, Beebe CA, Brunzell JD, Chi- serum lipid profiles in patients with type 2 diabetes. asson JE, Garg A, Holzmeister LA, Hoogwerf B, Diabetes Obes Metab. 2004: 6: 133–56. Mayer-Davis E, Mooradian AD, Purnell JQ and 93. Chui PC, Guari HP, Lehrke M and Lazar MA. Wheeler M. Evidence-based nutrition principles and PPAR-gamma regulates adipocytes cholesterol recommendations for the treatment and prevention metabolism via oxidized LDL receptor 1. J Clin of diabetes and related complications. Diabetes Care Invest. 2005; 15: 2244–56. 2002; 25: 148–98. 94. Ericsson CG, Hamsten A, Nilsson J, Grip L. Svane B 84. Cook S, Weitzman M, Auinger P, Nguyen M and and de Faire U. Angiographic assessment of effects of Dietz WH. Prevalence of a metabolic syndrome phe- bezafibrate on progression of coronary artery disease notype in adolescents: findings from the third National in young male post infarction patients. Lancet 1996; Health and Nutrition Examination Survey, 1988-1994. 347: 849–53. Arch Pediatr Adolesc Med. 2003; 157: 821–7. 95. Forman BM, Chen J and Evans RM. Hypolipidemic 85. Lakka TA, Laaksonen DE, Lakka HM, Mannikko N, drugs, polyunsaturated fatty acids and eicosanoids are Niskanen LK, Rauramaa R and Salonen JT. Seden- ligands for peroxisome proliferator-activated recep- tary lifestyle, poor cardiorespiratory fitness, and the tors alpha and delta. Proc Natl Acad Sci USA 1997; metabolic syndrome. Med Sci Sports Exerc. 2003; 35: 94: 4312–17. 1279–86. 96. Guerin M, Bruckert E, Dolphin PJ, Turpin G and 86. National Cholesterol Education Program (NCEP) Ex- Chapman MJ. Fenofibrate reduces plasma cholesteryl pert Panel on Detection, Evaluation. and Treatment ester transfer from HDL to VLDL and normalizes the of High Blood Cholesterol in Adults (Adult Treat- atherogenic, dense LDL profile in combined hyper- ment Panel III). Third Report of the National Choles- lipidemia. Arterioscler Thromb Vasc Biol. 1996; 16: terol Education Program (NCEP) Expert Panel on 763–72. Detection, Evaluation. and Treatment of High Blood 97. Van Gaal LF, Rissanen A, Scheen A, Ziegler O and Cholesterol in Adults (Adult Treatment Panel III) final Rossner S. Effect of rimonabant on weight reduction report. Circulation 2002; 106: 3143–421. and cardiovascular risk. Lancet 2005; 366: 369–70. 87. Grundy SM, Abate N and Chandalia M. Diet com- 98. Miettinen TA, Pyorala K, Oisson AG, Musliner TA, position and the metabolic syndrome: what is the Cook TJ, Faergeman O, Berg K, Pedersen T and optimal fat intake? Am J Med. 2002: 113 (Suppl. 9B): Kjekshus J. Cholesterol-lowering therapy in women 25S–29S. and elderly patients myocardial infarction or angina 88. Klein S, Sheard NF, Pi-Sunyer X, Daly A, pectoris: findings from the Scandinavian Simvas- Wylie-Rosett J, Kulkarni K and Clark NG. Weight tatin Survival Study (4S). Circulation 1997; 96: management through lifestyle modification for the 4211–18. 12.9 CONCLUSIONS 181

99. Colquhoun D, Keech A, Hunt D, Marschner I, Simes 104. Chobanian AV, Bakris GL, Black HR, Cushman WC, J, Glasziou P, White H, Barter P, Tonkin A and Green LA, Izzo JL Jr, Jones DW, Materson BJ, Oparil the LIP1D Study Investigators. Effects of pravastatin S, Wright JT Jr and Roccella EJ. Seventh Report of the on coronary events in 2073 patients with low lev- Joint National Committee on Prevention, Detection, els of both low-density lipoprotein cholesterol and Evaluation, and Treatment of High Blood Pressure. high-density lipoprotein cholesterol: results from the Hypertension 2003; 42: 1206–52. LIPID study. Eur Heart J. 2004; 25: 771–7. 105. Julius S, Kjeldsen SE, Brunner H. Hansson L, Platt 100. Stein DT, Devaraj S, Balis D, Adams-Huet B and F, Ekman S, Laragh JH, McInnes G, Schork AM, Jialal I. Effect of statin therapy on remnant lipopro- Smith B, Weber M and Zanchetti A. The VALUE trial: tein cholesterol levels in patients with combined Long-term blood pressure trends in 13,449 patients hyperlipidemia. Arterioscler Thromb Vasc Biol. 2001; with hypertension and high cardiovascular risk. Am J 21: 2026–31. Hypertens. 2003; 16: 544–8. 101. Parhofer KG, Barrett PH and Schwandt P. Atorvas- 106. Colwell JA. Antiplatelet agents for the prevention tatin improves postprandial lipoprotein metabolism in of cardiovascular disease in diabetes mellitus. Am J normolipidemic subjects. J Clin Endocrinol Metab. Cardiovasc Drugs 2004; 4: 87–106. 2000; 85: 4224–30. 107. Pearson TA, Blair SN, Daniels SR, Eckel RH, Fair 102. Paolisso G, Barbagallo M, Petrella G, Ragno E, Barbi- JM, Fortmann SP, Franklin BA, Goldstein LB, Green- eri M, Giordano M and Varricchio M. Effects of sim- land P, Grundy SM, Hong Y, Miller NH, Lauer RM, vastatin and atorvastatin administration on insulin re- Ockene IS, Sacco RL, Sallis JF Jr, Smith SC Jr, Stone sistance and respiratory quotient in aged dyslipidemic NJ and Taubert KA. AHA Guidelines for Primary Pre- non-insulin dependent diabetic patients. Atherosclero- vention of Cardiovascular Disease and Stroke: 2002 sis 2000; 150: 121–7. Update: Consensus Panel Guide to Comprehensive 103. Pontrelli L, Parris W, Adeli K and Cheung RC. Ator- Risk Reduction for Adult Patients Without Coronary vastatin treatment beneﬁcially alters the lipoprotein or Other Atherosclerotic Vascular Diseases. American proﬁle and increases low-density lipoprotein particle Heart Association Science Advisory and Coordinating diameter in patients with combined dyslipidemia and Committee. Circulation 2002; 106: 388–91. impaired fasting glucose type 2 diabetes. Metabolism 2002: 51: 334–42. 13 Perspectives on Diabetes Care in Old Age: A Focus on Frailty

Andrej Zeyfang1 and Jeremy D. Walston2 1Bethesda Hospital Stuttgart, Department of Internal Medicine and Geriatrics, Stuttgart, Germany 2Johns Hopkins University School of Medicine, Baltimore, MD, USA

including mobility disability, incontinence, depression Key messages and frailty [1]. These syndromes, when combined with diabetes and related medical complications, represent • Glucose intolerance and diabetes likely lead an especially great challenge to the quality of life of to a decline in muscle strength, which in turn these older patients. Older diabetic patients with these influences frailty. syndromes also present a major challenge to health care • The biology that underlies frailty – namely, the providers because of the complexity of their cases, and activation of inflammatory pathways and neu- because no clear evidence-based treatment regimens roendocrine dysregulation – influences a num- have been developed for this subset of diabetics. This ber of common chronic disease states in older adults. is in part because evidence-based guidelines aimed at • Diabetes and geriatric syndromes (immobility, younger individuals suffering from diabetes cannot be depression, dementia) are interdependent and simply extrapolated to the elderly without considering influence themselves reciprocally. Therefore, aspects such as multimorbidity, functional impairment, frailty develops more often in geriatric patients frailty and the need to evaluate individual goals, in- with diabetes. cluding life perspectives and the quality of life. For • The early identification of geriatric syndromes this rapidly growing group of older diabetics, there is a in older diabetics through a comprehensive large demand for data to be acquired from randomized geriatric assessment and an organized care plan controlled trials that seek the best and safest methods can improve the quality of life. to treat diabetes and its complications. The quest is also ongoing for data that focus on reducing the period of disability and increasing the quality of life. 13.1 Introduction Frailty is an important issue in geriatric medicine, as it characterizes that subset of older adults who are most Today, the prevalence of diabetes in older adults is in- vulnerable to adverse health care outcomes. This syn- creasing rapidly [1]. In addition to the common medical drome is relevant for the diabetologist who cares for an complications of diabetes, such as cardiovascular dis- ever-growing number of frail, older patients because ease (CVD), stroke and renal failure, older adults are of the extreme vulnerability exhibited by this subset much more susceptible to late-life geriatric syndromes, of patients to adverse health outcomes and iatrogenic

Diabetes in Old Age Third Edition Edited by Alan J. Sinclair © 2009 John Wiley & Sons Ltd. ISBN: 978-0-470-06562-4 184 CH 13 PERSPECTIVES ON DIABETES CARE IN OLD AGE: A FOCUS ON FRAILTY complications. In this chapter we will review the defini- Table 13.1 Components of different frailty screening tions and consequences of frailty, and also summarize tools. the biology of frailty, as understood to date. A review Chin 1999 [6] Studenski 2003 Fried 2001 [2] of the interactions between diabetes and frailty, and [4] modal pathways of how diabetes may accelerate frailty • inactivity and how frailty may accelerate or worsen diabetes, will • gait speed • slow walking be provided. Finally, a discussion is included of how • weight loss • ability to rise speed frailty may impact upon the treatment of diabetes and from a chair • poor hand grip other geriatric syndromes. • balance skills • exhaustion • 13.2 The frailty syndrome and its weight loss • low energy expen- biology diture • Frailty has long been recognized by practitioners of (three out of five) medicine as a syndrome of late-life, multisystem decline associated with vulnerability to adverse health outcomes, including accelerated mortality [2]. How- levels of insulin-like growth factor-1 (IGF-1) and de- ever, until very recently few investigators have at- hydroepiandrosterone sulphate (DHEA-S) (hormones tempted to identify its underlying aetiology and how which are important in muscle mass maintenance); and it might interact with chronic diseases such as dia- between high levels of cortisol and frailty [8–11]. Other betes. This is in part due to the heterogeneity with investigators have identified low levels of vitamin D as which frailty and many complex diseases of late life a strong correlate of frailty [12], and lower levels of present. Recently, several investigators have attempted red blood cells [13]. Importantly (as discussed below), to operationalize and characterize some of the clinical hyperglycaemia and hyperinsulinaemia are strongly re- and biological characteristics of frailty (see Table 13.1). lated to frailty [14, 15]. The apparent involvement of The definition proposed by Chin [6] is easiest to uti- multiple systems has led investigators to hypothesize lize because it has the fewest measurements. However, weight loss tends to be less predictive of adverse out- that frailty represents a multisystem decline in regula- comes when inserted into other frailty models, and tory and metabolic systems, and that this decline mani- hence it is a relatively weak predictor. The Studenski fests itself in declines in strength and energy as well as model [4] offers utility in that it is all based on objec- in a frank vulnerability to decline and death [16, 17]. tive measurements, but it has not been widely tested in other populations. The Fried model [2] is currently the most widely utilized in both outcomes and biolog- 13.3 Frailty and diabetes in late life ical research; this takes longer to complete because it incorporates two questionnaires, but these two criteria Many studies have previously demonstrated an in- are more subjective. Most emerging definitions related creased prevalence between diabetes and late-life syn- to frailty have focused on the concept of weakness, dromes, including frailty. For example, in the Cardio- fatigue, low levels of activity and the accumulation of vascular Health Study (CHS), a study of the evolution deficits and incorporate performance measures into the of CVD in over 5000 adults aged over 65 years, dia- examination [2–7]. betes and many other medical conditions were found The frailty metric outlined by Fried and cowork- to be significantly over-represented in the frail com- ers, which includes measures of fatigue, grip strength, pared to non-frail subsets of the cohort (Figure 13.1) walking speed, weight loss and activity levels, has [2]. The study results showed that 25% of the subset been instrumental in helping other investigators ex- of that population deemed frail was diabetic, 18.2% of plore and characterize the biology that underlies frailty the pre-frail or intermediate subset was diabetic, while [2]. These studies have identified significant relation- only 12% of the non-frail subset was diabetic. Fur- ships between frailty status and the inflammatory me- thermore, within the same population, it was apparent diators interleukin-6 (IL-6), C-reactive protein (CRP) that those frail CHS participants who were not diabetic and white blood cell counts; between frailty and low were significantly more likely to have higher glucose 13.4 THE BIOLOGY OF LATE-LIFE DIABETES 185

60 adults suggest that fasting hepatic glucose production is not increased [19]. Instead, there appears to be 50 two metabolically distinct groups of older diabetics: (i) a lean subset that is insulin-deﬁcient; and (ii) 40 an obese subset that is mostly insulin-resistant with 30 irregularities in glucose-stimulated pulsatile insulin secretion [20]. The older, insulin-deﬁcient subset 20 may be less able to respond with an appropriate

Proportion of patients (%) 10 glucose-triggered insulin release, in part because of an increased apoptosis of pancreatic beta cells associated 0 with both ageing and with increased fat mass and Heart insufficiency Cancer Diabetes Hypertension insulin resistance [21]. 14% 16.4% 32.4% 48.5% There are likely several aetiologies for the age-related increases in insulin resistance represented Figure 13.1 Prevalence of frailty in different medical conditions [14]. by the second group of diabetics. First, age-related declines in skeletal muscle, combined with increased fat mass, is likely to be a major contributing factor and insulin levels at baseline and after any oral glu- to insulin resistance in older adults [22]. Second, cose tolerance test than those who were not frail [14]. age-related alterations in energy expenditure and A more recent study in the same cohort has helped generation, and the level of the mitochondria likely to demonstrate that insulin resistance per se predicts contribute to abnormalities in glucose utilization incident frailty, providing important evidence that di- and energy expenditure and hence insulin resistance abetes and its underlying biology in part may drive [23]. Next – although not definitively demonstrated frailty [15]. Other supportive evidence of the relation- in vivo – increasing numbers of senescent cells likely ship between frailty and diabetes derives from another emerge with age. In-vitro models have suggested cohort study of older adults which showed that diabetes that these cells secrete inflammatory cytokines, accelerates the loss of skeletal muscle strength – an which in turn contributes to insulin resistance via the important component of frailty [18]. down-regulation of adiponectin, a potent regulator Despite the fact that these studies have indicated the of glucose uptake in adipose cells and a potent existence of a strong biological relationship between anti-inflammatory agent [24, 25]. Finally, multiple diabetes and frailty, or at least the strength component gene variants, including those in mitochondrial DNA, of frailty, many biological questions remain to be are known to influence the development of insulin answered. Regardless of the fact that there is some resistance [26]. Many of these variants likely remain evidence that incident frailty develops in the face of subclinical or undetectable until later in life, when diabetes, it may also be true that the biology that other age-related changes in body composition or underlies frailty may also drive the development of environmental changes, such as decreased activity, diabetes, or at least insulin resistance. In the following leads to the expression of the phenotype. In addition sections we will outline the biology that may contribute to the multiple age-related biological aetiologies to late-life diabetes, and describe how diabetes, its of insulin resistance, there are also important complications and its biology may influence frailty. In environmental and disease-related aetiologies that the final section we will suggest how frailty and its impact insulin resistance with aging. First, multiple underlying biology may drive diabetes. chronic disease states trigger chronic inflammatory responses, which in turn exacerbate insulin resistance in older adults [27]. Further, corticosteroids given to 13.4 The biology of late-life diabetes treat chronic inflammatory conditions worsen insulin resistance. Chronic disease states and functional Although younger and middle-aged individuals who decline leads to decreased activity levels, which in develop diabetes generally have multiple underlying turn contributes to insulin resistance. In summary, one metabolic alterations, including increased hepatic dominant biological aetiology of insulin resistance gluconeogenesis, increased insulin resistance and related to aging per se likely does not predominate decreased insulin secretion, data acquired from older in any given older diabetic. Rather, it is likely that 186 CH 13 PERSPECTIVES ON DIABETES CARE IN OLD AGE: A FOCUS ON FRAILTY multiple aetiologies coexist in each individual with controlled diabetes, which in turn negatively impacts insulin resistance, and that many of these age-related skeletal muscle. Furthermore, chronic inflammation is abnormalities increase with age. also known to negatively influence skeletal muscle strength in older adults, and is also known to be 13.4.1 How might diabetes and its strongly related to incident frailty [15]. This associ- underlying biology impact frailty? ation may be due to the accumulation of age-related glycosylated end products (AGE) which, when bound Although it is clear that diabetes and frailty are closely to specific receptors (RAGE, a signal transduction re- associated in clinical studies of older adults, to date ceptor), leads to the activation of nuclear factor-kappa only minimal evidence has been acquired from lon- B(NFκB)-related inflammatory pathways, which in gitudinal studies which suggests that one condition turn leads to the chronic secretion of inflammatory cy- is causal for the other. Rather, given the biologi- tokines such as IL-6; this causes a reinforcement of cal aetiologies described above for both syndromes, glucose intolerance as well as an acceleration of mus- it is highly likely that diabetes impacts the devel- cle mass decline [31]. Finally, it is also clear that a opment of frailty, and that frailty and its underlying number of diabetes complications, such as renal fail- biology contributes to the hyperglycaemia and/or dia- ure and CVD, can also activate inflammatory pathways betes frequently observed in frail, older adults. When and influence multiple systems into decline [32]. attempting to answer the question of how might diabetes impact frailty, several potential pathophys- 13.4.2 How might frailty impact the iological pathways are possible (Figure 13.2). First development of diabetes? and foremost, skeletal muscle appears to be negatively impacted by diabetes and glucose intolerance Frailty is likely driven by multiple age-related [18, 28, 29]. As skeletal muscle weakness is a key molecular and physiological changes, many of which component of frailty, it is highly likely that skele- are known to impact impaired glucose intolerance tal muscle weakness drives much of the association and diabetes (Figure 13.3). At a molecular level, between frailty and diabetes. The term ‘sarcopenic obe- age-related declines in mitochondrial function sity’ was coined to help describe the phenomena of are driven by damaged mitochondrial DNA. This older, obese, insulin-resistant or diabetic individuals cumulative change can result in lower levels of energy who have marked muscle weakness consistent with production, an impaired utilization of energy, and frailty [30]. The muscle decline in this subgroup of the increased production of free radicals of oxygen ‘frail’ older adults is likely profoundly influenced by [23]. The lower levels of energy production result both the fatty replacement of muscle tissue and by in glucose intolerance and lower metabolic rates in the insulin resistance, increased levels of cytokines, tissues such as skeletal muscle that normally have high and increased levels of adiponectin observed in this energy requirements. The increased generation of free population [30]. Additional pathophysiology related radicals associated with mitochondrial dysfunction to diabetes may also influence frailty. Chronic fluxes leads directly to the activation of NFκB-related in glucose levels related to diabetes appear to acti- inflammatory pathways. Chronic activity of this vate both the sympathetic nervous system (SNS) and pathway results in the chronic elevation of the the hypothalamic-pituitary-adrenal axis (HPA). The re- pleotrophic cytokine IL-6, the elevation of which in sult of the chronic activation of stress pathways is turn has been shown to influence the development of the chronic elevation of cortisol in those with poorly glucose intolerance, which in susceptible individuals

Activation of inflammation FRAILTY Diabetes Sarcopenia

HPA and SNS dysregulation

Figure 13.2 Schematic of key intermediate biology that connects diabetes to frailty. HPA = hypothalamic-pituitary-adrenal axis; SNS = sympathetic nervous system. 13.5 GERIATRIC SYNDROMES AND DIABETES 187

Molecular Physiological Clinical

Mitochondrial HPA axis Glucose-intolerant abnormalities dysregulation Diabetes Senescent cells Decreased energy utilization Muscle weakness Accelerated apoptosis of islet Glucose intolerance Frailty cells Decreased insulin secretion

Figure 13.3 Modal pathway between age-related molecular changes, altered physiology and clinical outcomes. HPA axis = hypothalamic-pituitary-adrenal axis. pushes towards the development of type 2 diabetes. In understand the biology that links the two, and to un- addition, there is considerable evidence that senescent derstand how best to treat older frail diabetics while cell types, including fibroblasts, endothelial cells and maintaining a high quality of life. Many investigators immune system cells, emerge with age. These cells have found that chronic disease, such as diabetes, poses appear to lose the ability to undergo apoptosis and a higher risk condition for the development of frailty develop inappropriate inflammatory characteristics. (see Figure 13.3) or functional decline [34]. In fact, The resultant increase in inflammatory mediators several studies have increasingly demonstrated the va- may further reinforces the inflammatory milieu that lidity of the hypothesis that the elderly with diabetes influences late-life glucose intolerance. Finally, ac- have a higher prevalence of geriatric syndromes. celerated apoptosis – or programmed cell death – has also been hypothesized to be accelerated in frailty and 13.5.1 Mobility, falls and urinary in other aging-related conditions such as Parkinson’s incontinence syndrome. Given the evidence that pancreatic beta-cell With increasing age, mobility begins to be affected in apoptosis is accelerated with ageing, it could be elderly individuals with diabetes. Park and coworkers speculated that declines in insulin production, in identified an accelerated loss of leg muscle strength and parallel with increased insulin resistance, may push quality in older adults with type 2 diabetes [18]. There frail individuals towards the development of diabetes. is also a higher risk of hip fracture [35]. Activities On a more physiological level, circulating inflam- in daily life are hindered, particularly those requiring matory mediators and increased cortisol secretion mobility, and incontinence may be problematic [36]. characterize frailty [5]. Chronic cortisol elevation Urinary incontinence is a common syndrome in the has been demonstrated to influence impaired glucose elderly with diabetes. In fact, its prevalence in women intolerance. Furthermore, chronic elevations in with impaired fasting glucose is almost double that in inflammatory mediators such as IL-6 trigger glucose women with a normal fasting glucose [37]. A typical tolerance and hence may influence the development concomitant problem is the need to urinate arising of diabetes in frail, older adults (Figure 13.3). suddenly (‘urge’), with falls frequently occurring as the subject tries to reach the toilet ‘in time’. 13.5 Geriatric syndromes and diabetes 13.5.2 Malnutrition Tooth loss and parodontitis are more prevalent in dia- In addition to the evidence for a strong relationship betes [38] and, together with problems in swallowing between diabetes and frailty, there is also evidence (e.g. after stroke), malnutrition and weight loss often for the mutual interaction between diabetes and geri- occur; these are important components of frailty. atric syndromes such as incontinence, immobility and In contrast to younger patients, in the elderly with cognitive or affective disorders [33]. Because of the diabetes undernutrition is more common; for this strong relationship between diabetes and multiple geri- reason, restrictive diets are not advantageous for atric syndromes, it will be increasingly important to elderly patients with diabetes and should not be 188 CH 13 PERSPECTIVES ON DIABETES CARE IN OLD AGE: A FOCUS ON FRAILTY prescribed. Careful attention should be paid to patients of frailty, diabetes and cognitive impairment. In other with weight loss (indicator of frailty) or tooth loss, as studies, a correlation has been identified between cog- functional pairs are very important for chewing. nitive dysfunction, as shown by a clock-drawing test or clock in a box and HbA1c [40]. 13.5.3 Depression Cognitive dysfunction can be worsened by a concomitant depression or by other drugs such as Late-life depression is found commonly even in benzodiazepines or pain medication [61]. Cognitive individuals without diabetes, reaching a prevalence of dysfunction may also render difficult all types of approximately 25%. Depression, reaching much higher self-management skills of the elderly. Hyperglycaemia rates in patients with diabetes, contributes to poor is associated with declines in cognitive function, and adherence to medication and dietary regimens. Depres- the extent of cognitive dysfunction in turn influences sion is the most important determinant for reduced ad- the ability to attain any quality of diabetes control. herence to medications [47] and causes therefore poor Thus, at times the goals of tight glucose control may glycaemic control. The depressed elderly have a re- need to be lowered. duced quality of life, and increased health care expenditures [39]. In younger adults, a history of depression 13.5.5 Interaction between geriatric increases the risk of diabetes [54]. Although depres- syndromes and diabetes sion is a risk factor for mortality in older patients with diabetes [55], the implementation of a depression The mutually influencing effects of coexisting dia- management programme would have a decreased betes and geriatric syndromes are a very important risk of mortality compared to depressed patients field of future research. Treatment goals and strategies with diabetes in usual-care practices [56]. Before depend heavily on functional resources and deficits elderly patients take part in educational programmes [36]. Examples of this interdependence are listed in for diabetes, depression should be screened for, and Table 13.2, and the respective chapters in this book pro- treated if indicated. vide a deeper insight into the pathophysiological basis of depression or dementia and their connections to di- 13.5.4 Cognition abetes. As most of these geriatric syndromes are either directly or indirectly components of the frailty syn- Cognitive dysfunction and dementia are also more drome, it is imperative that professionals in this field common in the diabetic older adult, and are in part cor- should carefully examine for the presence of functional related with diabetes control. This link was described problems in elderly patients with diabetes. by Reaven et al. in 1990 [57], but no explanation was provided at the time regarding the reasons. In the descriptive Rotterdam-Study [58] there was a dou- 13.6 Approaches for improving care ble prevalence of vascular or degenerative dementia for frail, older diabetics in patients suffering from diabetes. Only during the past few years have any studies provided any insight With knowledge of these interactions and the impact into the particular way in which diabetes might be on quality of life, new strategies for the comprehensive linked to cognitive dysfunction. Type 2 diabetes is care of the geriatric patient with diabetes are needed. strongly associated with cognitive impairment, especially in semantic memory and perceptual speed, but 13.6.1 Comprehensive geriatric assessment not with episodic memory, working memory or visuospatial ability, nor with any measure of global cog- Before frailty develops, the resources and deficits of nition [59]. Importantly, metabolic syndrome is also the geriatric patient should be assessed and quantified significantly associated with declines in all cognitive by performing a comprehensive geriatric assessment. measures. Among the single components of metabolic The effectiveness in reducing short-term mortality, the syndrome, hyperglycaemia was most strongly and sig- increased ability to live at home for at least one year, nificantly associated with cognitive function in a recent and improvement of physical and cognitive function, study [60]. The same study highlighted a significant has been proven for those living in the community interaction between metabolic syndrome and inflam- [41] as well as for hospitalized patients [42]. Some mation on cognition, indicating the close connection recommendations for the selection of elderly patients 13.7 DRUG THERAPY IN THE FRAIL ELDERLY 189

Table 13.2 Interdependence of diabetes and geriatric syndromes [43]. Syndrome Effect on diabetes Diabetes causes Dementia • Worsens HbA 1c • Prevalence of dementia higher • Diabetes education and self-management ren- • Worse metabolic control causes more severe cog- dered difﬁcult nitive impairment • Controls obfuscated (retinopathy, blood-glucose and -pressure, foot, feeding and drinking)

• • Depression Worsens HbA1c Prevalence of depression higher • Reduced compliance • Depressive feeling augmented by fear of seque- • ‘Pseudodementia’ lae • Feeling guilty about ‘diet errors’

Incontinence • Voluntary reduced drinking amount causing • With hyperglycaemia, a large amount of urinary dehydration and hyperglycaemia ﬂow • Urinary tract infections causing worse metabolic • High prevalence of urge incontinence control • End-stage, often overﬂow, incontinence

Immobility • Physical activity more difficult • Dizziness or fatigue hinders activity • Self-care difficult or impossible (foot-care, blood • Polyneuropathies cause afferent ataxia glucose self-control, insulin injection) • Falls and hip fractures more often in diabetes with diabetes who should undergo comprehensive geri- falls and fractures, enhanced cognitive function, and atric assessment are listed in Table 13.3, together with generally the prevention of disease and frailty [45]. some clues regarding the use of assessment tools. 13.6.4 Nutrition 13.6.2 Educational programmes It is common practice today to advise the diabetic It has been perceived that conventional education pro- patient on diet; however, given the prevalence of grammes for diabetes patients are not as effective for undernutrition in the geriatric age group, therapeutic the elderly; consequently, new approaches have been regimens calling for restrictive diets should clearly designed for application in elderly patients [43]. Apart be stopped for these patients. Nutritional recom- from a particular didactic approach, age-specific sub- mendations to geriatric patients should be realistic, jects are also required. The general recommendations individualized, and acceptable to the patient. Not include individualized approaches, risk–benefit consid- only does food play an important role in the quality erations, and the involvement of multiple disciplines of life, but their body weight is also crucial when and care partners [44]. Moreover. special educational considering morbidity and mortality. A loss of body programmes for care-givers in an out-patient setting or weight past the age of 75 years can be dangerous; a −2 nursing home are needed. body mass index (BMI) <22.7 kg m in old age is associated with above-average mortality rates, even in 13.6.3 Physical training the apparently healthy [46]. The classic approach of encouraging more physical activity for the younger patient with diabetes is 13.7 Drug therapy in the frail most likely not workable for the elderly diabetic. elderly Instead, low-threshold, slower approaches are encouraged which motivate the patient, combine the The effects, side effects and particularities of drug positive effects of physical activity on metabolism therapy in the elderly with diabetes in general are with the benefits on muscular strength, reduced risk of described elsewhere; here, attention is focused on the 190 CH 13 PERSPECTIVES ON DIABETES CARE IN OLD AGE: A FOCUS ON FRAILTY

Table 13.3 Recommendations for selection of elderly patients with diabetes that should undergo comprehensive geriatric assessment [52] and recommended assessment tools [53]. Presence of a geriatric syndrome Presence of several coexisting Presence of disabilities resulting from morbidities, apart from diabetes lower-limb vascular disease or neuropathy requiring a rehabilitation programme Confused state MMSE Complex drug More than ﬁve Diabetic foot CDT regimens different drugs syndrome CAM

Depression GDS PNP Peripheral Falls TuG More than two drugs neuropathy 5-Chair-Rise for the same (Semmes– Modiﬁed Romberg test condition Weinstein- (Standing positions) monoﬁlament More than once-daily 10 g) administration of a drug Immobility TCS PAD Ankle-brachial Pressure sores Braden-Scale index (Doppler) Incontinence Malnutrition MNA NRS In the absence of a terminal illness or severe dementia syndrome

MMSE = Mini-Mental-State-Examination; CDT = clock-drawing test; CAM = Confusion Assessment Method; GDS = Geriatric Depres- sion Scale; TuG = Timed Up & Go-Test; TCS = Trunk Control Scale; MNA = Mini-Nutritional Assessment; NRS = Nutritional Risk Score; PNP = peripheral neuropathy; PAD = peripheral arterial disease

administration of the medication. Adherence to therapy to determine the individual’s ability to self-administer (either insulin or oral drugs) is generally low, ranging insulin correctly, the cognitive, visual and fine motor from 61 to 85% in a 6-month period of observation capacities of the patient should be examined. Here, for for oral medication to approximately 63% with in- example, the timed money-counting test could be used sulin therapy [47]. Nikolaus et al. showed that, in a [51]. Before the patient is provided with an insulin group of elderly patients, 10% failed to open at least pen, the visus, cognitive capacity, ability to correctly one container of medication given in different dosage self-dose and the obstacles to insulin self-application forms. This inability was associated with poor vision, should be examined. It is necessary to carefully assess impaired cognitive function and low manual dexterity. the benefits of self-management (‘I can eat whenever Compliance with prescribed medication correlated to I want’) versus dependency (‘I have to wait until the cognitive function, ability to handle medication con- nurse has time for me’). It should also be determined tainers, number of prescribed drugs, and recent changes if the elderly patient requires professional support for in drug prescriptions [48]. insulin therapy. As of yet, no insulin-injecting device specifically aimed for the elderly has been developed, 13.7.1 What about insulin? but this would be helpful. Insulin therapy poses a major challenge in the treat- 13.7.2 Assistive devices ment of geriatric patients. Insulin is not only a potent agent to normalize blood sugar, but it also acts as an There is a substantial demand for innovative assis- anabolic drug, allowing ‘doping’ of the frail elderly tive devices for the elderly with diabetes. This begins [49]. Unfortunately, the administration of insulin is with age-appropriate devices for self-control of the problematic for the elderly (see Table 13.4), as well blood sugar level. The obstacles when using existing as achieving the appropriate dosage – which if not cor- devices are multiple: too small; non-user friendly; a rect will lead to hypo- or hyperglycaemia [50]. In order need for calibration; digital numbers that could be 13.8 SUMMARY 191

Table 13.4 Difficulties in self-administration of insulin the incidence of hip-fracture after a fall. The preven- in the elderly. tion of falling is a priority and is achievable; as many falls occur in the bathroom, handles or grab-bars are Forgetting injection (no dosage) Forgetting that already injected (double dosage) advisable, installing a movement-sensor to turn on the Wrong dosage (‘U for 0’ – 50 Units given, prescription light or putting a toilet-chair next the bed are simple 5U) to do. These basic procedures can facilitate remaining No mixing before injection of biphasic or NPH-Insulin at home versus an unwanted move into a nursing No change of needle – obstruction home. For hospitalized patients with diabetes, socks Not waiting the appropriate time interval after injecting with small rubber buttons to prevent slipping could be insulin Waiting too long after self-administered medication worn at night to avoid the need for putting on shoes. (taking a nap) Sensor-mats that signal an alarm when the patient leaves the bed at night are useful for hospitalized patients, especially when at risk for disorientation or misinterpreted if read upside-down, and so on. Intelli- delirium. The aims of achieving a more comfortable, gent blood glucose-monitoring systems for the elderly safe and acceptable existence for the elderly patient should be much easier to use. An advanced but sim- with diabetes, as well as extending the length of time ple system capable of providing advice that is easy to living without dependence on others, can be attained understand would be an example. For instance, simi- by prudent operation of assistive devices. lar to a traffic light where green indicates ‘everything is fine’, orange might mean ‘take care’, and red ‘call 13.8 Summary for medical help’. Modern electronic and wireless technology could be included, quietly transmitting data, for In summary, it is increasingly clear that there are im- example, to a nursing centre. portant interactions between diabetes, frailty and other Devices used for insulin injection today are not geriatric syndromes in older adults. These interactions optimal for the elderly. These should be constructed appear to have a deeply biological basis that im- so that aspects such as limited motor ability or pact profoundly upon skeletal muscle and are, in part, tremor, reduced visual function and diminished driven by neuroendocrine and inflammatory changes. strength of the thumb (when using a pen) are Most importantly, as frail, older adults are more vulner- considered. State-of-the-art technology could also able to adverse health outcomes, it is critical to be able be used in a device to enable the insulin-delivery to identify the frailest subset of diabetic patients and system to provide the desirable dose for the moment to formulate a clear and coherent treatment plan that automatically (an expert system might be connected to focuses on the quality of life and reduction of symp- the glucose-monitoring system). Clearly, adaptations toms in these most vulnerable patients. The interaction to the design of insulin devices would be one means of geriatric syndromes – mainly depression, dementia, of enabling the elderly to begin insulin therapy malnutrition and diabetes – should, on the other hand, with more ease, comfort and acceptance, but further always be considered in the treatment plan. Multidis- innovative studies are required to help facilitate the ciplinary approaches targeting increased functionality delivery of insulin and increase safety in this process. and independent living can be effective in this regard, When treating frail elderly with diabetes it is im- considering also the particularities and possibilities in portant to understand other basic assistive devices and drug therapy and the use of assistive devices. their limitations. It is useful to test visual and hearing capacities by screening to identify a need for glasses or hearing aids, particularly before starting an educational References programme. The prevention of falls and hip-fracture is also part of the care of the diabetic elderly. Canes, 1. Blaum C. (2008) Descriptive epidemiology of diabetes. walkers or rolling-walkers are not only useful for In: M. Munshi and L. Lipsitz (eds). Geriatric Diabetes individuals at high risk for falling (e.g. because of Book. Taylor and Francis Group, LLC, New York, neuropathy and afferent ataxia), but can also be pp. 1–10. temporarily of help, for example when wearing a cast 2. Fried LP, Tangen C, Walston J, Newman A, Hirsch CH, for a diabetic foot. Wearing hip-protectors can reduce Gottdiener JS, Seeman T, Tracy R, Kop WJ, Burke G 192 CH 13 PERSPECTIVES ON DIABETES CARE IN OLD AGE: A FOCUS ON FRAILTY

and McBurnie MA. Frailty in older adults: evidence 14. Walston J, McBurnie MA, Newman A, Tracy R, Kop for a phenotype. Journal of Gerontology 2001; 56A WJ, Hirsch CH, Gottdiener JS and Fried LP. Frailty and (3): M1–11. activation of the inflammation and coagulation systems 3. Rockwood K and Mitnitski A. Frailty in relation to the with and without clinical morbidities: Results from the accumulation of deficits. J Gerontol A Biol Sci Med Cardiovascular Health Study. Arch Intern Med 2002; Sci 2007; 62 (7): 722–7. 162: 2333–41. 4. Studenski S, Hayes RP, Leibowitz RQ, Bode R, Lavery 15. Barzilay JI, Blaum C, Moore T, Xue QL, Hirsch CH, L, Walston J, Duncan P and Perera S. Clinical Global Walston JD and Fried LP. Insulin resistance and in- Impression of Change in Physical Frailty: development flammation as precursors of frailty: the Cardiovascular of a measure based on clinical judgment. J Am Geriatr Health Study. Arch Intern Med 2007; 167 (7): 635–41. Soc 2004; 52 (9): 1560–6. 16. Walston J. Frailty – the search for underlying causes. 5. Walston J, Hadley EC, Ferrucci L, Guralnik JM, New- Sci Aging Knowledge Environ 2004; (4): e4. man AB, Studenski SA, Ershler WB, Harris T and 17. Fried LP, Hadley EC, Walston JD, Newman A, Gu- Fried LP. Research agenda for frailty in older adults: ralnik JM, Studenski S, Harris TB, Ershler WB and toward a better understanding of physiology and eti- Ferrucci L. From bedside to bench: research agenda ology: summary from the American Geriatrics Soci- for frailty. Sci Aging Knowledge Environ 2005; (31): ety/National Institute on Aging Research Conference e24. on Frailty in Older Adults. J Am Geriatr Soc 2006; 54 18. Park SW, Goodpaster BH, Strotmeyer ES, Kuller LH, (6): 991–1001. Broudeau R, Kammerer C, de RN, Harris TB, Schwartz 6. Chin APM, Dekker JM, Feskens EJ, Schouten EG and AV, Tylavsky FA, Cho YW and Newman AB. Accel- Kromhout D. How to select a frail elderly population? erated loss of skeletal muscle strength in older adults A comparison of three working definitions. J Clin with type 2 diabetes: the health, aging, and body com- Epidemiol 1999; 52 (11): 1015–21. position study. Diabetes Care 2007; 30 (6): 1507–12. 7. Studenski S, Perera S, Wallace D, Chandler JM, Dun- 19. Meneilly GS and Elahi D. Metabolic alterations in can PW, Rooney E, Fox M and Guralnik JM. Physical middle-aged and elderly lean patients with type 2 performance measures in the clinical setting. J Am diabetes. Diabetes Care 2005; 28 (6): 1498–9. Geriatr Soc 2003; 51 (3): 314–22. 20. Meneilly GS, Veldhuis JD and Elahi D. Deconvolution 8. Leng S, Chaves P, Koenig K and Walston J. Serum analysis of rapid insulin pulses before and after six interleukin-6 and hemoglobin as physiological corre- weeks of continuous subcutaneous administration of lates in the geriatric syndrome of frailty: a pilot study. glucagon-like peptide-1 in elderly patients with type J Am Geriatr Soc 2002; 50 (7): 1268–71. 2 diabetes. J Clin Endocrinol Metab 2005; 90 (11): 9. Leng SX, Cappola AR, Andersen RE, Blackman MR, Koenig K, Blair M and Walston JD. Serum 6251–6. levels of insulin-like growth factor-I (IGF-I) and 21. Fridlyand LE and Philipson LH. Reactive species, dehydroepiandrosterone sulfate (DHEA-S), and their cellular repair and risk factors in the onset of type 2 relationships with serum interleukin-6, in the geriatric diabetes mellitus: review and hypothesis. Curr Diabetes syndrome of frailty. Aging Clin Exp Res 2004; 16 (2): Rev 2006; 2 (2): 241–59. 153–7. 22. Hughes VA, Roubenoff R, Wood M, Frontera WR, 10. Varadhan R, Walston J, Cappola AR, Carlson MC, Evans WJ and Fiatarone Singh MA. Anthropometric Wand GS and Fried LP. Higher levels and blunted assessment of 10-y changes in body composition in diurnal variation of cortisol in frail older women. Journal the elderly. Am J Clin Nutr 2004; 80 (2): 475–82. of Gerontology: Medical Science 2008, 63 (2), 190–5. 23. Figueiredo PA, Mota MP, Appell HJ and Duarte JA. 11. Leng SX, Xue QL, Tian J, Walston J and Fried LP. The role of mitochondria in aging of skeletal muscle. Inflammation and frailty in older women. Journal of Biogerontology 2008, 9(2), 67–84. the American Geriatrics Society 2007; 55 (6): 864–71. 24. Guerre-Millo M. Adiponectin: An update. Diabetes 12. Puts MT, Visser M, Twisk JW, Deeg DJ and Lips P. Metab 2008, 34 (1), 12–18. Endocrine and inflammatory markers as predictors of 25. Tilg H and Moschen AR. Role of adiponectin and frailty. Clin Endocrinol (Oxf) 2005; 63 (4): 403–11. PBEF/visfatin as regulators of inflammation: involve- 13. Chaves PH, Semba RD, Leng SX, Woodman RC, ment in obesity-associated diseases. Clin Sci (Lond) Ferrucci L, Guralnik JM and Fried LP. Impact of 2008; 114 (4): 275–88. anemia and cardiovascular disease on frailty status 26. Walston J and Silver K. (2008) The genetics of diabetes of community-dwelling older women: the Women’s and its complications in older adults. In: M. Munshi and Health and Aging Studies I and II. J Gerontol A Biol L. Lipsitz (eds). Geriatric Diabetes Book.Taylorand Sci Med Sci 2005; 60 (6): 729–35. Francis Group, LLC, New York, pp. 11–28. 13.8 SUMMARY 193

27. Libby P. Inflammatory mechanisms: the molecular ba- dysfunction is associated with poor diabetes control in sis of inflammation and disease. Nutr Rev 2007; 65 older adults. Diabetes Care 2006; 29 (8): 1794–9. (12Pt 2): S140–6. 41. Stuck AE, Egger M, Hammer A, Minder CE and Beck 28. Goodpaster BH, Carlson CL, Visser M, Kelley DE, JC. Home visits to prevent nursing home admission and Scherzinger A, Harris TB, Stamm E and Newman functional decline in elderly people: systematic review AB. Attenuation of skeletal muscle and strength in the and meta-regression analysis. JAMA 2002; 287 (8): elderly: The Health ABC Study. J Appl Physiol 2001; 1022–8. 90 (6): 2157–65. 42. Gordon JN, Trebble TM, Ellis RD, Duncan HD, Johns 29. Park SW, Goodpaster BH, Strotmeyer ES, de RN, T and Goggin PM. Thalidomide in the treatment of Harris TB, Schwartz AV, Tylavsky FA and Newman cancer cachexia: a randomised placebo controlled trial. AB. Decreased muscle strength and quality in older Gut 2005; 54 (4): 540–5. adults with type 2 diabetes: the health, aging, and body 43. Zeyfang A. [Structured educational programs for geri- composition study. Diabetes 2006; 55 (6): 1813–18. atric patients with diabetes mellitus]. MMW Fortschr 30. Roubenoff R. Sarcopenic obesity: does muscle loss Med 2005; 147 (26): 43, 45-3, 46. cause fat gain? Lessons from rheumatoid arthritis and 44. Suhl E and Bonsignore P. Diabetes self-management osteoarthritis. Ann N Y Acad Sci 2000; 904: 553–7. education for older adults: general principles and prac- 31. Ershler WB and Keller ET. Age-associated increased tical application. Diabetes Spectrum 2006; 19: 234–40. interleukin-6 gene expression, late-life diseases, and 45. Warburton DE, Nicol CW and Bredin SS. Health ben- frailty. Annu Rev Med 2000; 51: 245–70. efits of physical activity: the evidence. Canadian Med- 32. Cesari M, Leeuwenburgh C, Lauretani F, Onder G, ical Association Journal 2006; 174 (6): 801–9. Bandinelli S, Maraldi C, Guralnik JM, Pahor M and 46. Breeze E, Clarke R, Shipley MJ, Marmot MG and Ferrucci L. Frailty syndrome and skeletal muscle: re- Fletcher AE. Cause-specific mortality in old age in sults from the Invecchiare in Chianti study. Am J Clin relation to body mass index in middle age and in old Nutr 2006; 83 (5): 1142–8. age: follow-up of the Whitehall cohort of male civil 33. Hader C, Beischer W, Braun A, Dreyer M, Friedl A, servants. Int J Epidemiol 2006; 35 (1): 169–78. Fusgen I, Gastes U, Grunklee D, Hauner H, Kobberling 47. Cramer JA. A systematic review of adherence with J, Kolb G, von Laue N, Muller UA and Zeyfang A. medications for diabetes. Diabetes Care 2004; 27 (5): Diagnosis treatment and follow up of diabetes mellitus 1218–24. in elderly. European Journal of Geriatrics 2006; 8: 1–57. 48. Nikolaus T, Kruse W, Bach M, Specht-Leible N, Oster 34. Gregg EW, Mangione CM, Cauley JA, Thompson TJ, P and Schlierf G. Elderly patients’ problems with Schwartz AV, Ensrud KE and Nevitt MC. Diabetes medication. An in-hospital and follow-up study. Eur and incidence of functional disability in older women. J Clin Pharmacol 1996; 49 (4): 255–9. Diabetes Care 2002; 25 (1): 61–7. 49. Sonksen PH. Insulin, growth hormone and sport. J 35. Lipscombe LL, Jamal SA, Booth GL and Hawker Endocrinol 2001; 170 (1): 13–25. GA. The risk of hip fractures in older individuals 50. Grissinger M. Top 10 adverse drug reactions and med- with diabetes: a population-based study. Diabetes Care ication errors. Program and abstracts of the American 2007; 30 (4): 835–41. Pharmacists Association Annual Meeting, 16 March 36. Zeyfang A. [Treating diabetes in the very old]. MMW 2007. Fortschr Med 2007; 149 (19): 29-3. 51. Nikolaus T, Bach M, Oster P and Schlierf G. The 37. Brown JS, Vittinghoff E, Lin F, Nyberg LM, Kusek Timed Test of Money Counting: a simple method JW and Kanaya AM. Prevalence and risk factors for of recognizing geriatric patients at risk for increased urinary incontinence in women with type 2 diabetes and health care. Aging (Milano) 1995; 7 (3): 179–83. impaired fasting glucose: findings from the National 52. Sinclair AJ. Special Considerations in older adults with Health and Nutrition Examination Survey (NHANES) diabetes: meeting the challenge. Diabetes Spectrum 2001-2002. Diabetes Care 2006; 29 (6): 1307–12. 2006; 19 (4): 229–33. 38. Presson SM, Niendorff WJ and Martin RF. Tooth loss 53. Zeyfang A. (2006) In: Diabetologie kompakt,4th and need for extractions in American Indian and Alaska edition. H. Schatz (ed.). Thieme, Stuttgart, Ch. 3.6, Native dental patients. J Public Health Dent 2000; 60 pp. 135–9. (Suppl. 1): 267–72. 54. Brown LC, Majumdar SR, Newman SC and Johnson JA. 39. Lustman PJ and Clouse RE. Depression in diabetic History of depression increases risk of type 2 diabetes in patients: the relationship between mood and glycemic younger adults. Diabetes Care 2005; 28: 1063–7. control. J Diabetes Complications 2005; 19 (2): 113–22. 55. Katon WJ, Rutter C, Simon G, Lin EH, Ludman E, 40. Munshi M, Grande L, Hayes M, Ayres D, Suhl E, Ciechanowski P, Kinder L, Young B and Von Korff M. Capelson R, Lin S, Milberg W and Weinger K. Cognitive The association of comorbid depression with mortality 194 CH 13 PERSPECTIVES ON DIABETES CARE IN OLD AGE: A FOCUS ON FRAILTY

in patients with type 2 diabetes. Diabetes Care 2005; dementia: The Rotterdam Study. Neurology 1999; 53 28: 2668–72. (9): 1937–42. 56. Bogner HR, Morales KH, Post EP and Bruce ML. Dia- 59. Arvanitakis Z, Wilson RS, Li Y, Aggarwal NT and betes, depression, and death. A randomized controlled Bennett DA. Diabetes and function in different cog- trial of a depression treatment program for older adults nitive systems in older individuals without dementia. based in primary care (PROSPECT). Diabetes Care Diabetes Care 2006; 29 (3): 560–5. 2007; 30: 3005–10. 60. Dik MG, Jonker C, Comijs HC, Deeg DJ, Kok A, 57. Reaven GM, LW Thompson, D Nahum and E Hask- Yaffe K and Penninx BW. Contribution of metabolic ins. Relationship between hyperglycemia and cognitive syndrome components to cognition in older individuals. function in older NIDDM patients. Diabetes Care 1990; Diabetes Care 2007; 30 (10): 2655–60 13: 16–21. 61. Nikolaus T and Zeyfang A. Pharmacological treatments 58. Ott A, Stolk RP, van Harskamp F, Pols HAP, Hofman for persistent non-malignant pain in older persons. A and Breteler MMB. Diabetes mellitus and the risk of Drugs Aging 2004; 21 (1): 19–41 14 Metabolic Decompensation in the Elderly

Giuseppe Paolisso and Michelangela Barbieri Department of Geriatric Medicine and Metabolic Diseases, Second University of Naples (SUN), Piazza Miraglia, Napoli, Italy

The attainment of glycaemic control in order to re- Key messages duce acute complications associated with diabetes is a priority for all elderly patients. Although ﬂexibil- • Both, diabetic ketoacidosis and hyperosmolar ity may be allowed in setting glycaemic parameters hyperglycaemic state may be the presenting for individual treatment goals, the clinician also must features in older people without a previous take into account the need to prevent associated acute diagnosis of diabetes. complications. • Early detection and prompt treatment of both Not only higher glycaemic values, but also even metabolic disturbances can minimize the excess more dangerous hypoglycaemic episodes, can disturb mortality seen in ageing patients. the frail equilibrium of older patients. Hence, achieving target glycaemic goals while avoiding hypogly- 14.1 Introduction caemia represents a major challenge in the management of elderly patients with diabetes mellitus. Re- peated episodes of hypoglycaemia may cause extreme Impaired glucose homeostasis plays an important role emotional distress in such patients, even when the in aged individuals, its manifestations varying from episodes are relatively mild. light – often unrecognized – hyperglycaemia up to acute hyperglycaemic crises that frequently lead to hospital admission. 14.2 Hypoglycaemia Hyperglycaemic complications include dehydration, mental status changes, increased risk of infection and, Hypoglycaemia accounts for a relatively high number in severe cases, the possibility of ketoacidosis and hy- of emergencies requiring hospital admission. The con- perosmolar coma, a condition which is more often fatal dition occurs far more frequently as a consequence in the elderly. In the elderly, whereas the mortality rate of an inadequate therapy in diabetes management, for diabetic emergencies associated with ketoacidosis as demonstrated in the Diabetes Control and Com- has remained low, that for diabetic emergencies as- plications Trial (DCCT) and the United Kingdom sociated with a hyperosmolar state has, by contrast, Prospective Diabetes Study (UKPDS) [1, 2]. The fre- remained considerably higher. quency of hypoglycaemic episodes increases with in-

Diabetes in Old Age Third Edition Edited by Alan J. Sinclair © 2009 John Wiley & Sons Ltd. ISBN: 978-0-470-06562-4 196 CH 14 METABOLIC DECOMPENSATION IN THE ELDERLY creasing quality of glycaemic control assessed with to hypoglycaemia. Early studies of the effect of age HbA1c. It is generally accepted that hypoglycaemia oc- on counter-regulatory hormone responses to hypogly- curs when the capillary blood glucose level is below caemia produced inconsistent results [10, 11], although 0.6gl−1. Symptomatic hypoglycaemia can be distin- these studies were seen to be flawed because many of guished from ‘silent’ hypoglycaemia, which is fre- the elderly subjects included also suffered from under- quently associated with the occurrence of severe hy- lying diseases. Recent investigations, in which older poglycaemia [1]. subjects free from disease have been carefully selected, In older patients, susceptibility to hypoglycaemia is have shown that healthy elderly subjects have im- pronounced, and is exacerbated by older people hav- paired glucagon responses to hypoglycaemia [12, 13]. ing little knowledge about the symptoms and signs of Epinephrine responses have been reported to be either the condition [3, 4]. While hypoglycaemic events are impaired [12] or increased [13] in healthy elderly; cor- usually of minor importance, in some instances death tisol responses have been found to be either normal or severe sequelae – including myocardial infarction [12] or increased [13], and GH responses may be im- and stroke – may occur. Hypoglycaemia in the elderly paired [12], although this has not been demonstrated subject can have serious, sometimes life-threatening, conclusively. consequences for the heart or brain, both in terms A reduced epinephrine response to hypoglycaemia of morbidity and quality of life. In some cases the and a decreased responsiveness to β-adrenergic recep- outcome may be permanent neurological damage, pre- tor stimulation may explain the reduced awareness of sumably because of an already compromised cerebral the autonomic symptoms of hypoglycaemia found in circulation [4]. the healthy elderly when compared to young subjects Consequently, in older subjects the risk of hypogly- [14]. caemia must be evaluated as clearly as possible and In young adults, symptomatic responses to hypogly- balanced, on an individual basis, against the potential caemia are generated at a blood glucose level that is benefit of a near-normal glucose level. higher than that at which cognitive function becomes The symptoms differ somewhat from those observed impaired. This allows sufficient time to take correc- in younger subjects (notably blurred vision and in- tive action before severe neuroglycopenia supervenes stability), and are often blunted by an autonomous [15]. The difference between these glycaemic thresh- neuropathy or impaired cognitive function (‘silent’ olds is ∼1.0 mmol l−1 (18mgdl−1). Indeed, in older hypoglycaemia). Hypoglycaemic symptoms in the el- persons the difference between the glycaemic threshold derly tend to present predominantly as neuroglycopenic for subjective awareness of hypoglycaemia and that for symptoms, including impaired concentration, personal- the onset of cognitive dysfunction may be absent [16]. ity changes, focal neurological deficits, seizure or syn- Thus, elderly patients who became hypoglycaemic are cope. Nocturnal hypoglycaemia may present as morn- less likely to experience prior warning symptoms if ing headaches or disturbed sleep. Adrenergic symp- their blood glucose level falls, and are at greater risk toms (tremulousness, anxiety, diaphoresis, palpitation for injury or falls with fracture – all factors which have and hunger) are diminished, in part due to a loss of a disruptive effect on the frail subject. autonomic nerve function [5]. However, two recent studies have described less Glucose counter-regulatory hormones such as serious consequences in elderly subjects provided with glucagon, epinephrine (adrenaline) and growth hor- adequate care. For example, mortality was zero in a mone (GH) are the most important hormones secreted study conducted in Germany [5], while morbidity was in response to hypoglycaemia. When glucagon is only 4% in a study from Singapore [3]. An additional deficient, epinephrine becomes critical, while GH point was that elderly diabetics were less exposed and cortisol are important if the hypoglycaemia to motor vehicle accidents [5], which is a serious is prolonged [6]. The elderly exhibit an impaired complication of hypoglycaemia observed in the young response of glucose counter-regulatory hormones in population of the DCCT [1]. the presence of decreased glucose. Furthermore, the Hypoglycaemia is usually observed during the late rate of insulin clearance from the circulation declines morning and afternoon in insulin-treated patients. The with age, which may lead to an enhancement of the favouring factors, other than age, include multiple co- risk of hypoglycaemia in elderly people [7–9]. morbid conditions (psychiatric conditions and depres- Changes in the release of counter-regulatory hor- sion leading to variable food intake), renal impair- mones may increase the susceptibility of the elderly ment (sulphonylureas cause a ninefold increase in the 14.2 HYPOGLYCAEMIA 197 risk of severe hypoglycaemia, especially if food in- and tolbutamide than with glibenclamide or chlor- take is irregular), multiple medications (high-risk as- propamide, while hypoglycaemia is fourfold more fre- sociation with antibacterial sulfamides) and, more frequent with glibenclamide than with gliclazide [17]. quently, a poorly adapted behavior response. To this Prolonged-release glipizide is associated with severe list should be added the rare use of self-monitoring hypoglycaemia in elderly subjects and/or in subjects and the absence of patient and caregiver education with renal impairment. Unlike conventional sulphony- regarding the symptoms of hypoglycaemia. Further- lureas, glinides taken before a meal induce a rapid more, the direct cause of hypoglycaemia is gener- postprandial insulin response. The short half-life of ally related to a dietary error (53% of hypoglycaemic these drugs ensures that insulin concentrations peak episodes follow a missed meal) and/or recent hospi- at 1–2 h, but by 6 h have returned to fasting concentrations with minimal risk of hypoglycaemia if the talization (change in therapy poorly adapted to home patient misses a meal; this, on the other hand, is a life) [4], African-American race, and the use of five severe problem with the older sulphonylureas. Sup- or more concomitant medications [17, 18]. The fre- port for such clinical evidence is derived from studies quency of severe hypoglycaemia remains moderate which showed the risk of severe hypoglycaemia to among patients with type 2 diabetes (0.4 episode per be less than half that seen with traditional sulphony- 100 patient-years), irrespective of treatment, compared lureas [23].The short half-life and biliary elimination with insulin-treated patients (1.5 episodes per 100 of glinides are interesting properties; however, as with patient-years). glitazones, specific large-scale studies in elderly per- Hypoglycaemia is a complication of insulin or sons are lacking. any drug that increases pancreatic insulin secretion In the future, it would be useful to evaluate the or release (sulphonylureas and glinides). Antidiabetic effects of rapid insulin analogues and glinides on the monotherapies that do not normally cause hypogly- hypoglycaemic risk in an elderly population aged over caemia include metformin, the alpha-glucosidase in- 70 years. hibitors and the thiazolidinediones. However, if these In conclusion, elderly type 2 diabetics suffer a higher therapies are added to insulin or drugs that promote frequency of hypoglycaemic episodes, with the highest insulin release, then the incidence of hypoglycaemia frequency and most severe hypoglycaemia being ob- may increase. served in insulin-treated patients. Particular at-risk sit- Hypoglycaemia is threefold more frequent when in- uations include chronic renal failure and polymedica- sulin is used alone compared to antidiabetic drugs, tion, both of which are frequently present in this popu- with the combined treatments exposing the patient to lation. The prescription of long-acting sulphonylureas, an intermediary risk. Recent advances in molecular especially glibenclamide, is not recommended in el- genetic engineering have made possible the develop- derly patients. Targeting ‘perfect’ glycemic control, ment of insulin analogues with pharmacokinetics that which is associated with a significant hypoglycaemic more closely mimic the needs of patients with type risk, is not justified in many elderly patients. In patients 2 diabetes, thus reducing the risk of hypoglycaemia. aged over 65 years, the recommended therapeutic targets for glycaemic control should be fixed at a higher Lys.Pro insulin and insulin aspart (both rapid-acting in- level than for younger patients. In order to prevent sulin analogues) administered immediately prior to the severe – if not fatal – hypoglycaemia in old patients, meal have demonstrated improved postprandial glu- a careful insulin scheduling (when necessary) should cose control in comparison with regular insulin. In- include preparations characterized by a shorter half-life jection just before a meal, or even during a meal, and with minimal risk for nocturnal episodes. The short represents a good way of adapting the dose to real half-life sulfamides, and agents tolerated in cases of food intake. However, in practice multiple injections impaired renal and hepatic function, should be the first are not always easy to implement in the elderly subject choice for geriatric care. Finally, due to the fact that [19–22]. elderly patients are often incapable of treating hypo- For oral antidiabetic drugs, α-glucosidases and met- glycaemia themselves, educational programmes should formin do not normally cause hypoglycaemia. The include advice and information relating to the detec- pharmacokinetic properties of sulfamides would favour tion and treatment of hypoglycaemia, including the the use of second-generation, short-acting drugs. The criteria for hospital admission in cases of unresponsive risk of severe hypoglycaemia is lower with glipizide hypoglycaemia [24]. 198 CH 14 METABOLIC DECOMPENSATION IN THE ELDERLY 14.3 Diabetic ketoacidosis and Table 14.1 The main features of diabetic ketoacidosis (DKA) and hyperosmolar hyperglycaemic state (HHS) in hyperosmolar hyperglycaemic older people. state in the elderly DKA Diabetic ketoacidosis (DKA) and hyperosmolar • Rare in the elderly non-ketotic coma (HONK) are two of the most • Patients less likely to be receiving insulin before devel- serious acute complications in the spectrum of marked oping DKA decompensated diabetes [25–28]. • These hyperglycaemic emergencies persist as impor- Patients less likely to have had a previous episode of DKA tant causes of morbidity and mortality among diabetic patients, despite major advances in the understand- • More insulin required to treat the DKA ing of their pathogenesis and more uniform agreement • A longer length of hospital stay about their diagnosis and treatment. In contrast to DKA • Higher mortality rate mortality, the mortality rate of hyperosmolar hyper- • glycaemic state (HHS) has remained high at ∼15%, No specific treatment guidelines available compared to <5% in patients with DKA [27, 29–31]. HHS Severe dehydration, older age and the presence of co- • morbid conditions in patients with HHS account for Occurs frequently in the elderly the higher mortality in these patients [31]. • In ∼50% of cases diabetes mellitus has not been previ- DKA consists of the biochemical triad of hyper- ously diagnosed or treated glycaemia, ketonaemia and metabolic high anion gap • Frequent predisposing factors are: acidosis [29]. • impaired maintenance of serum osmolality; The terms ‘hyperglycaemic hyperosmolar non- • decreased thirst perception (especially in elderly with ketotic coma’ and ‘hyperglycaemic hyperosmolar dementia); non-ketotic state’ have been replaced with the term • decreased access to water, especially in the bed-ridden ‘hyperglycaemic hyperosmolar state’ in order to and use of diuretics; reflect the facts that: (i) alterations of the sensoria • acute infection (pneumonia being the most common may often be present without coma; and (ii) the infection). hyperosmolar hyperglycaemic state may consist of • Symptoms, signs, diagnosis, and treatment are otherwise moderate to variable degrees of clinical ketosis [26]. similar to those in younger adults. Although DKA most often occurs in patients with type 1 diabetes mellitus (T1DM), it may also occur in type 2 diabetes under conditions of extreme stress DKA, typically require more insulin to treat the DKA, such as serious infection, trauma, cardiovascular or have a longer length of hospital stay, and have a higher other emergencies. Less often, it will present as a mortality rate (22% for those aged ≥ 65 years versus manifestation of type 2 diabetes, a disorder known as 2% for those aged <65) [36]. Causes of death include ‘ketosis-prone type 2 diabetes’ [30]. Similarly, whereas infection, thromboembolism and myocardial infarction HHS occurs most commonly in T2DM, it can be seen [36]. Although concomitant diseases and high rates of in T1DM in conjunction with DKA [32, 33]. morbidity must be considered when caring for older Both, DKA and HHS are characterized by absolute patients with DKA, no specific treatment guidelines or relative insulinopenia. Clinically, they differ only are currently available. by the severity of dehydration, ketosis and metabolic HHS almost always occurs in older people, and in acidosis [34, 35]. While both DKA and HHS can be about half of the cases diabetes mellitus has not been seen in the elderly [36], DKA is rare and its features previously diagnosed or treated [38]. The predisposi- and management do not differ from those in younger tion of the elderly to develop HHS can be explained diabetics. However, its mortality is greatest in old by a combination of impaired maintenance of serum age, particularly because of associated cardiovascular osmolality, a decreased thirst perception (especially in disease (CVD) [36, 37] (Table 14.1). Older patients are elderly with dementia), a decreased access to water less likely to be receiving insulin before they develop (especially in the bed-ridden) and the use of diuret- DKA, are less likely to have had a previous episode of ics. A reduced thirst perception renders the polydipsia 14.4 PATHOGENESIS OF DKA AND HHS 199 less dramatic, thereby lessening recognition by self or breakdown of triglycerides and the release of free fatty others, leading to dehydration and ending up in hy- acids (FFAs). The FFAs are taken up by the liver and perosmolar coma [39]. An acute infection is the most converted to ketone bodies that are released into the frequent predisposing factor (40–60%), with pneumo- circulation. This process of ketogenesis is stimulated by nia being the most common infection. Other illnesses the increase in glucagon levels [40]. Glucagon will ac- such as stroke, acute myocardial infarction, renal insuf- tivate carnitine palmitoyltransferase I, an enzyme that ficiency and medications such as glucocorticoids can allows FFAs in the form of coenzyme A to cross the also be predisposing factors. mitochondrial membrane following their esterification The symptoms, signs, diagnosis and treatment are into carnitine. On the other hand, esterification is re- otherwise similar to those in younger adults. versed by carnitine palmitoyltransferase II to form fatty acyl coenzyme A, which enters the β-oxidative pathway to produce acetyl coenzyme A. Most of the latter 14.4 Pathogenesis of DKA and HHS is utilized in the synthesis of β-hydroxybutyric acid and acetoacetic acid – two relatively strong acids that In both DKA and HHS, the underlying metabolic ab- are responsible for the acidosis in DKA. normality results from the combination of absolute or Normally, ketone bodies increase insulin release relative insulin deficiency and increased amounts of from the pancreas, and the insulin in turn suppresses counter-regulatory hormones. Inadequate levels of cir- ketogenesis. In the insulin-deficient state, however, culating insulin lead to hyperglycaemia, which in turn the pancreatic β cells are unable to respond and the can lead to progressive dehydration and hyperosmolar- ketogenesis will proceed unchecked. ity and ultimately to HHS. If the insulin deficiency is The reason for the absence of ketosis in the presence severe enough, ketosis and ultimately acidosis will de- of insulin deficiency in HHS remains unknown [26]. velop. A relative insulin deficiency – not an absolute The current hypothesis is that the absence may be due insulin deficiency – is necessary for the development to the lower levels of FFAs or the higher portal vein of both DKA and HHS. Even patients with T2DM and insulin levels, or both [25, 41, 42]. It appears that ‘normal insulin levels’ may develop DKA if the level in hyperglycaemic coma, most subjects with T2DM of insulin resistance causes a sufficiently large increase have just enough residual insulin secretion to suppress in insulin requirement. lipolysis and ketogenesis, thus avoiding DKA and When insulin is deficient, the elevated levels of glucagon, catecholamines and cortisol will stimu- developing a HONK coma instead. However, in one late hepatic glucose production through increased study similar insulin levels were found in subjects with glycogenolysis and enhanced gluconeogenesis DKA or HONK, whereas those with HONK had lower [29]. Hypercortisolaemia will result in increased levels of counter-regulatory hormones, leading to less proteolysis, thus providing amino acid precursors for lipid breakdown and less hepatic ketogenesis [43]. It gluconeogenesis. also appears that hyperosmolality not only worsens Low insulin and high catecholamine concentrations insulin resistance but also inhibits lipolysis [44]. will reduce glucose uptake by the peripheral tissues. The combination of an elevated hepatic glucose pro- 14.4.1 Acid–base balance, fluids and duction and a decreased peripheral glucose utilization electrolytes is the main pathogenic disorder responsible for hyperglycaemia in DKA and HHS. The hyperglycaemia will Acidosis in DKA is due to the overproduction of lead to glycosuria, osmotic diuresis and dehydration. β-hydroxybutyric acid and acetoacetic acid. At physio- Initially, glycosuria causes an increase in the glomeru- logical pH, these two ketoacids dissociate completely, lar filtration rate (GFR), but when the hypovolaemia and the excess hydrogen ions bind the bicarbonate, re- becomes significant the GFR is decreased and renal sulting in decreased serum bicarbonate levels. Ketone glucose losses may also decrease. As glucose clear- bodies thus circulate in the anionic form, which leads ance by the kidney declines, the hyperglycaemia and to the development of anion gap acidosis that char- hyperosmolarity worsen. acterizes DKA. Despite substantial losses of ketoacids In DKA, the low insulin levels, combined with in the urine, the decrease in serum bicarbonate con- increased levels of catecholamines, cortisol and GH centration and increase in the anion gap observed in will activate hormone-sensitive lipase; this causes the DKA are almost equal [45]. Metabolic acidosis will 200 CH 14 METABOLIC DECOMPENSATION IN THE ELDERLY

induce hyperventilation through a stimulation of pe- diuresis, for a deficit of 1–2 mmol kg−1 on average ripheral chemoreceptors and the respiratory centre in [27, 41]. the brainstem, which will elicit a decrease in the partial pressure of carbon dioxide. This will partially compensate for the metabolic acidosis. 14.4.2 Precipitating factors Hyperglycemia-induced osmotic diuresis results in Infection remains the most important precipitating fac- severe fluid loss. The total body deficit of water is tor in the development of DKA and HHS. In 20–25% usually about 5–7 l in DKA and 7–12 l in HHS, which of cases, infections are the first manifestations of pre- represents a loss of about 10–15% of body weight. viously undiagnosed diabetes mellitus [48]. Omissions The osmotic diuresis is associated with large losses of or inadequate insulin doses are frequent precipitating electrolytes in the urine. The sodium chloride deficit factors, particularly for DKA [41]. − in DKA and HHS is usually 5–13 mmol kg 1 body Other precipitating factors, especially for HHS, are − weight for sodium, and 3–7 mmol kg 1 for chloride silent myocardial infarction, cerebrovascular accident, [26, 29, 41]. Initially, the increased glucose concentra- mesenteric ischaemia, acute pancreatitis and the use tion is restricted to the extracellular space, which forces of medications such as steroids, thiazide diuretics, water from the intracellular to the extracellular com- calcium-channel blockers, propranolol and phenytoin partment and induces a dilution of the plasma sodium [26]. In 2–10% of cases of DKA, no obvious precipi- concentration. Subsequently, further increases in the tating factor can be identified [48]. plasma glucose concentration will lead to osmotic diuresis, with losses of water and sodium chloride in the urine; the water loss usually exceeds that of the sodium 14.5 Diagnosis of DKA and HHS chloride [29, 45]. Because of the osmotic shift of water, plasma 14.5.1 Clinical presentation sodium concentrations are usually low or normal in DKA, but may be slightly increased in HHS, de- If a physical examination reveals dehydration along spite extensive water loss [45, 46]. In this context, with a high capillary blood glucose level, with or with- the plasma sodium concentration should be corrected out urine or increased plasma ketone bodies, then acute for hyperglycaemia by adding 1.6 mmol to the reported diabetic decompensation should be strongly suspected. sodium level for every 5.6 mmol l−1 increase in glucose A definitive diagnosis of DKA or HHS must be con- above 5.6 mmol l−1 [29]. The plasma sodium concen- firmed through laboratory investigations; however, the tration may also be artificially lowered by the presence clinical presentation can provide helpful information of severe hyperlipidaemia. for the preliminary bedside diagnosis [49]. Both, DKA and HHS are also associated with pro- DKA usually occurs in younger, lean patients with found total body potassium depletion, ranging from 3 T1DM and develops within a day or so, whereas to 15 mmol kg−1 of body weight [27, 41, 47]. However, HHS is more likely to occur in elderly, obese diabetic plasma potassium concentrations are typically normal patients, often those with decreased renal function who or elevated at the time of presentation. As with sodium, do not have access to water; in these cases the HHS the presence of hyperglycaemia leads to a shift of water may take days or weeks to fully develop [50]. and potassium from the intracellular to the extracellu- The pathophysiological consequences of hypergly- lar space. The shift of potassium is further enhanced in caemia, hyperketonaemia and insulin deficiency ac- the presence of acidosis, intracellular proteolysis and count for many of the classic symptoms and physical insulinopenia [48]. Potassium depletion is due to ex- findings seen in DKA and HHS. High glucose levels cessive urinary potassium loss secondary to osmotic lead to an osmotic diuresis, dehydration and ultimately diuresis, and leads to an increased delivery of fluid hypotension. The high ketone concentrations are re- and sodium to the potassium secretory sites in the dis- sponsible for the metabolic acidosis and also cause an tal nephron [29]. This can be further exacerbated by a osmotic diuresis. poor oral intake of potassium, vomiting and secondary Both, DKA and HHS often present with polyuria hyperaldosteronism [29]. and polydypsia, although polydypsia may be absent Phosphate, magnesium and calcium are other in elderly patients with HHS. In both conditions, elements excreted in excess in urine during the abdominal pain with nausea and vomiting can develop development of DKA and HHS owing to osmotic owing to acidosis per se or to decreased mesenteric 14.6 TREATMENT OF DKA AND HHS 201 perfusion, and can be mistaken for an acute surgical measure blood β-hydroxybutyric acid levels at the bed- abdomen. Kussmaul–Kien respiration (rapid and deep side, using a reagent strip and a reflectance meter [54]. respiration) with acetone on the breath is typical of The majority of patients presenting with DKA and DKA but is absent in HHS. Although dehydration oc- HHS have an elevated leukocyte count, usually in the − curs in both conditions, it is often more pronounced in range of 10.0–15.0 × 109 l 1, even in the absence HHS. Because DKA and HHS are usually accompanied of infection [25], this being attributed to stress and by hypothermia, a normal or elevated temperature may dehydration [48]. Amylase levels are often elevated in indicate an underlying infection. patients with DKA, but represent enzyme activity from Although patients may be alert at the time of presen- non-pancreatic tissues such as the parotid gland. Lipase tation, changes in their mental status are common and levels will usually be normal. Additional laboratory vary from confusion or disorientation to coma, usually tests should include blood culture, urinalysis and urine as a result of extreme dehydration with or without pre- culture, chest radiography and electrocardiography, as renal azotaemia, hyperglycaemia and hyperosmolarity. well as measurement of the lactate level, if indicated. In contrast to DKA, focal or generalized seizures and Because a high fetal mortality rate is associated with transient hemiplegia may occur. ketoacidosis, it is important to eliminate the possibility of pregnancy in women of reproductive age.

14.5.2 Laboratory findings 14.6 Treatment of DKA and HHS Most patients presenting with DKA have a plasma glucose level of 14 mmol l−1, or greater. However, most The therapeutic goals for the treatment of hypergly- patients with T1DM who have such a plasma glucose caemic crises in diabetes consist of: level do not have ketoacidosis. On the other hand, ketoacidosis may develop in patients with a plasma • Improving the circulatory volume and tissue perfu- glucose level below 14 mmol l−1. In HHS, hypergly- sion. caemia is usually more severe than in DKA, and a • Decreasing the serum glucose and plasma osmolality ≥ −1 plasma glucose level 34 mmol l is arbitrarily one towards normal levels. of the diagnostic criteria. Glucose is the main osmole responsible for the hyperosmolar syndrome. The in- • Clearing the serum and urine of ketones at a steady creased serum osmolality can be calculated as follows: rate. × + [(2 serum Na) serum glucose], with normal val- • Correcting electrolyte imbalances. ues being 290 ± 5 (SD) mmol kg−1 water. Blood urea nitrogen is not included in the calculation of effective • Identifying and treating any precipitating events osmolality because it is freely permeable in and out of [55]. the intracellular compartment [26, 51]. By definition, The successful treatment of DKA and HHS depends −1 the osmolality must exceed 320 mmol kg to be diag- on the adequate correction of dehydration, hypergly- nostic of HHS. However, it is not uncommon in DKA caemia, ketoacidosis and electrolyte deficits [56]. to have increased osmolality. In DKA the blood pH Any comorbid precipitating event should be identi- will be ≤ 7.3, and in HHS in isolation it will be >7.3. fied and treated appropriately. Both, DKA and HHS Venous blood can be used to measure pH and bicar- are medical emergencies, and patients with these con- bonate levels, unless information on oxygen transport ditions must be admitted to hospital. is required. It must be remembered that venous blood, The treatment of HHS involves frequent and careful without arterial blood gas values, does not permit the monitoring (Figure 14.1). Although 4–6 l of fluid may identification of mixed acid–base disorders [52]. In be needed during the first 12 h, such rapid replacement DKA, a lower pH will usually be associated with a may not be feasible in older persons, who often exhibit − decrease in bicarbonate to ≤ 15 mmol l 1, although a a poor cardiac reserve [57]. In most cases, insulin and milder form of DKA may present with a bicarbonate intravenous fluids can be safely started simultaneously, level of 15–18 mmol l−1. Less severe DKA is always the exceptions being patients with hypokalaemia or accompanied by moderate to large amounts of ketones hypotension. In such cases, intravenous fluids should in the blood and urine, while trace amounts may also be given before insulin to prevent any worsening be found in cases of HHS [53]. Today, it is possible to of hypokalaemia or hypotension that may occur in 202 CH 14 METABOLIC DECOMPENSATION IN THE ELDERLY

After history and physical examination, obtain complete blood count, chemistry profile and blood gases

IV fluids Insulin Potassium Bicarbonate

IV fluids should be done cautiously in light of Decrease plasma glucose at rate of Must be maintained in normal range to −1 −1 concerns in elderly patients of known or undiagnosed 50-75 mg dl h avoid cardiac complication. heart disease.The induced change in serum osmolality −1 −1 should not exceed 3 mOsm H2O kg h HHS or DKA DKA pH ≥ 7.0 pH < 7.0 Consider CVP and urinary catheter Regular insulin at dose of 0.1 U kg−1body weight IV bolus. none 44 mmol + HCO + If hypokalaemia (K 3 − −1 15 mEq KCl Isotonic saline 0.9% NaC l 15-20 ml kg 1 <3.3 m Eql ): hold insulin and 2 h until during the 1st hour in the absence of − + give 40 m Eqh 1 K pH > 7.0 cardiac compromise −1 −1 + − 0.1 U kg h as continuous until K >3.3m Eql 1 infusion until glucose <250 mg dl−1 Subsequent choice of fluid replacement depends on state of hydration, serum electrolytes and urine output: If hyperkalaemia: Check serum glucose hourly. If serum glucose does Correction of acidosis, − not fall by at least 50 mg dl 1 in first hour, then double Volume replacement, Insulin therapy − insulin dose hourly until glucose falls at a steady 1 − 0.45% NaCl if Na > 150-155 mmol l hourly rate of 50-70 mg dl 1 Do not give K+ but check potassium every 2 h Change to 5% dextrose + 0.45% NaCl − when serum glucose reaches 300 mg dl 1

CORRECTION OF CORRECTION OF CORRECTION OF ELECTROLYTE DEHYDRATION HYPERGLYCAEMIA AND IMBALANCE KETOACIDOSIS

Figure 14.1 Management of patients with DKA and HHS.

response to insulin, and the resulting intracellular shift first 24 h. It is important that the change in osmolality of glucose, potassium and water [58]. does not exceed 3 mmol kg−1 h−1 [26, 27, 46, 59, 60]. In patients with kidney and heart problems, their cardiac, renal and mental statuses must be assessed 14.6.1 Fluid therapy frequently, with regular serum osmolality monitoring The objective of an initial fluid therapy is to expand during rehydration to avoid iatrogenic water overload the extracellular volume (intravascular and extravascu- [26, 27, 46, 59, 60]. lar) and restore renal perfusion. In the absence of any Caution is indicated in elderly patients with heart major heart problems, it is suggested that treatment be failure or renal insufficiency in order to avoid fluid started with an infusion of isotonic saline (0.9% NaCl) overload. Along with frequent clinical and laboratory − − assessment, bladder catheterization and the monitor- at a rate of 15–20 ml kg 1 h 1 during the first hour ing of central venous pressure or pulmonary capillary (1–1.5 l in an average adult) so as to rapidly expand wedge pressure may be warranted in order to assess the extracellular space. The subsequent choice of fluid fluid status more accurately. replacement depends on the state of hydration, the electrolyte levels and urinary output. In general, this may be an infusion of 0.45% NaCl at a rate of 4–14 ml kg−1 14.6.2 Insulin therapy − h 1 if the serum sodium level is normal or elevated. There is general consensus that, in cases of DKA The administration of hypotonic saline leads to a grad- and HHS, regular insulin should be administered by ual replacement of the intracellular and extracellular means of continuous intravenous infusion in small compartments. As soon as the renal function is assured, doses through an infusion pump [29, 40, 59, 61]. potassium must be added to every litre of fluid. When Such low-dose insulin therapy provides insulin the plasma glucose level reaches 12–14 mmol l−1, each concentrations that are more physiological and litre of fluid should contain 5% dextrose. Fluid replace- produce a more gradual and steady fall in plasma ment should correct the estimated water deficit over the glucose levels [62, 63]; the risk of hypoglycaemia and 14.6 TREATMENT OF DKA AND HHS 203 hypokalaemia is also decreased [29]. The available agents may not be effective in patients with severe data do not support the subcutaneous or intramuscular fluid depletion as they are given subcutaneously. route for insulin administration [29]. Although most proposed protocols suggest that a loading dose of 14.6.3 Potassium therapy insulin should be given at the initiation of insulin therapy, there are no data to support any advantage The treatment of DKA and HHS with rehydration for such a recommendation [29, 63]. and insulin is typically associated with a rapid de- As soon as hypokalaemia (potassium concentration cline in the plasma potassium concentration, particu- <3.3 mmol l−1) has been excluded, a continuous in- larly during the first few hours of therapy [45, 48]. fusion of regular insulin can be started at a dose of This rapid decrease is due to several factors, the most 0.1 U kg−1 h−1, which should produce a gradual de- significant being the insulin-mediated re-entry of potas- crease in the plasma glucose level of 3–4 mmol l−1 sium into the intracellular compartment. Other factors h−1 [62]. When the plasma glucose level reaches are extracellular fluid volume expansion, correction 12–14 mmol l−1, the insulin infusion rate may be de- of acidosis, and continued potassium loss owing to creased by 50% as the 5% dextrose is added. There- osmotic diuresis and ketonuria. Despite major potas- after, the insulin infusion dose must be adjusted to sium depletion in the whole body, mild to moderate maintain the plasma glucose values until the acidosis hyperkalaemia is not uncommon in patients in hy- in DKA or the clouded consciousness and hyperosmo- perglycaemic decompensation. Because treatment will lality in HHS have been resolved. rapidly induce decreased serum potassium concen- When the ketoacidosis in DKA has been corrected trations, potassium replacement must be initiated as −1 (plasma glucose level <11.0 mmol l−1, serum bicar- soon as levels fall below 5.0 mmol l , assuming that bonate level ≥18 mmol l−1, venous pH >7.3 and anion the urine output is adequate. It is recommended that gap <12 mmol l−1), the clouded consciousness and hy- 20–30 mmol of potassium be added to each litre of perosmolality in HHS have resolved, and patients are infusion fluid to maintain the serum potassium con- −1 able to take fluids orally, then a multidose insulin reg- centration at 4–5 mmol l [56]. If the serum potassium −1 imen may be initiated based on the patient’s treatment level is less than 3.3 mmol l , then potassium replace- before the DKA or HHS developed. ment therapy should be started immediately with fluid Recent clinical studies have demonstrated the therapy, and the initiation of insulin therapy should be potency and cost-effectiveness of subcutaneous delayed until the potassium concentration is restored −1 rapid-acting insulin analogues (lispro or aspart) in to above 3.3 mmol l , in order to avoid arrhythmia, the management of patients with uncomplicated mild cardiac arrest and respiratory muscle weakness [56]. to moderate DKA [64, 65]. The patients received Initially, the serum potassium level should be mea- subcutaneous rapid-acting insulin doses of 0.2 U kg−1 sured every 1–2 h because the most rapid change oc- initially, followed by 0.1 U kg−1 every hour, or an curs during the first 5 h of treatment. Subsequently, it initial dose of 0.3 U kg−1 followed by 0.2 U kg−1 every should be measured every 4–6 h, as indicated clinically. 2 h until the blood glucose level was <250 mg dl−1 (13.7 mmol l−1). The insulin dose was then decreased 14.6.4 Bicarbonate and phosphate therapy by half to 0.05 or 0.1 U kg−1, respectively, and ad- Bicarbonate therapy ministered every 1 or 2 h until resolution of the DKA. No differences in the duration of hospital stay, total The use of bicarbonate in the treatment of DKA re- amount of insulin needed for the resolution of hyper- mains controversial [66]. Most current reviews do not glycaemia or ketoacidosis, or in the incidence of hy- recommend the routine use of alkali therapy in DKA, poglycaemia among treatment groups, were found [64, because the condition tends to correct with insulin 65]. The use of insulin analogues allowed the treatment therapy. Insulin administration inhibits ongoing lipol- of DKA in general wards or the emergency department, ysis and ketoacid production and promotes ketoanion and so reduced the cost of hospitalization by 30%, metabolism. Because protons are consumed during ke- without any significant changes in hypoglycaemic toanion metabolism, bicarbonate is regenerated, lead- events [65]. It is important to note here that the use of ing to a partial correction of any metabolic acidosis. fast-acting insulin analogues is not recommended for The rationale for bicarbonate therapy is the (theoreti- patients with severe DKA or HHS, as there no studies cal) assumption that severe acidosis could contribute to have been conducted to support their use. Again, these organ malfunction, such as of the liver, heart and brain. 204 CH 14 METABOLIC DECOMPENSATION IN THE ELDERLY However, few prospective, randomized studies of the all of which may be minimized by careful monitoring. use of bicarbonate in DKA have been conducted. Hyperchloraemia is a common, but transient, finding Studies of patients with a blood pH of 6.9 or higher that usually requires no special treatment. have found no evidence that bicarbonate is beneficial Cerebral oedema is a rare but important complica- [34], and some studies have even suggested that bi- tion of DKA. Although it can affect adults, it is more carbonate therapy might be harmful for these patients common in young patients. The early signs of cerebral [35–37]. Because no studies have been conducted in oedema include headache, confusion and lethargy, patients with a blood pH below 6.9, the administration but papilloedema, hypertension, hyperpyrexia and of bicarbonate as an isotonic solution is still recom- diabetes insipidus may also occur. Patients typically mended. However, it should be noted that, as bicar- improve mentally with an initial treatment of DKA, bonate therapy lowers the potassium levels, these will but then suddenly worsen. Multiple factors in the need to be monitored very carefully [67]. treatment of DKA and HHS may contribute to the cerebral oedema, including: (i) the idiogenic osmoles, Phosphate therapy which cannot be dissipated rapidly during rehydration, thus creating a gradient and a shift of water into The beneficial effect of phosphate therapy is purely the cells [29]; (ii) insulin therapy per se, which may theoretical. It would be expected to prevent poten- promote the entry of osmotically active particles into tial complications associated with hypophosphataemia, the intracellular space; and (iii) the rapid replacement such as respiratory depression, skeletal muscle weak- of sodium deficits [27, 47]. ness, haemolytic anaemia and cardiac dysfunction. Fur- In order to reduce the risk of cerebral oedema, it thermore, the majority of controlled, randomized trials is recommended that physicians correct sodium and have been unable to demonstrate any clinical benefit water deficits gradually, and avoid any rapid decline of routine phosphate therapy [29], but still recommend in plasma glucose concentration [25, 56]. that one-third of potassium replacement be given as potassium phosphate. No studies have been conducted 14.7.1 Adult respiratory distress syndrome on the use of phosphate therapy for HHS. Adult respiratory distress syndrome, or non- cardiogenic pulmonary oedema, is a potentially 14.6.5 Clinical and laboratory follow-up fatal complication of DKA that fortunately occurs rarely [29]. The partial pressure of oxygen, which is Vital signs should be monitored at 30-min intervals normal on admission, decreases progressively during for the first hour, hourly for the next 4 h, and then ev- treatment to unexpectedly low levels. This change is ery 2–4 h until resolution of the condition. An accurate believed to be due to increased water in the lungs record of hourly urine output is necessary to monitor and reduced lung compliance. These changes may be kidney function. On admission, a comprehensive pro- similar to those occurring in brain cells leading to file will include at least arterial or venous blood gas cerebral oedema, which suggests that it is a common values, levels of plasma glucose, electrolytes, blood biological phenomenon in tissues [29]. urea nitrogen and creatinine, ketone levels in the serum or urine (or both), and serum osmolality. Capillary 14.7.2 Hypochloraemic metabolic acidosis blood glucose levels should be monitored hourly to allow any adjustment of the insulin infusion dose. Elec- This phenomenon is not uncommon during the treat- trolyte levels should be measured every 1–2 h initially, ment of DKA [43]. A major mechanism is the loss of and every 4 h thereafter. The measurement of venous substrates (ketoanions) in the urine that are necessary pH can replace that of arterial pH, and should be un- for bicarbonate regeneration [43, 68]. Other mecha- dertaken every 4 h until the DKA has been corrected. nisms include: (i) intravenous fluids containing chloride concentrations exceeding that of plasma [68, 69]; (ii) volume expansion with bicarbonate-free fluids [68, 14.7 Treatment-related 69]; and (iii) an intracellular shift of sodium bicarbon- complications ate during the correction of DKA [70]. This acidosis usually has no adverse effect and is corrected sponta- Common complications of DKA include hypogly- neously in the subsequent 24–48 h through enhanced caemia, hypokalaemia and recurrent hyperglycaemia, renal acid excretion [68–71]. 14.8 CONCLUSIONS 205 14.7.3 Vascular thrombosis 5. Holstein A, Plaschke A and Egberts EH. Clinical char- acterisation of severe hypoglycaemia: a prospective Many features of DKA and HHS predispose the patient population-based study. Exp Clin Endocrinol Diabetes to thrombosis, including dehydration and contracted 2003; 111: 364–9. vascular volume, a low cardiac output, an increased 6. Cryer PE. Glucose counterregulation in man. Diabetes blood viscosity and the frequent presence of under- 1981; 30: 261–4. lying atherosclerosis [40, 51]. In addition, a number 7. Reaven GN, Greenfield MS, Mondon CE, Rosenthal of haemostatic changes favour thrombosis [72]. This M Wright D and Reaven EP. Does insulin removal complication is more likely to occur when the osmo- rate from plasma decline with age? Diabetes 1982; 31: lality is very high. Low-dose or low-molecular-weight 670–3. heparin therapy should be considered for prophylaxis 8. Minaker KL, Rowe JW, Torino R and Pallotta JA. In- in patients at high risk of thrombosis, although no data fluence of age on clearance of insulin in man. Diabetes are yet available demonstrating the safety or efficacy 1982; 31: 851–5. 9. Fink RI, Revers RR, Kolterman OG and Olefsky JM. of this approach. The metabolic clearance of insulin and the feedback inhibition of insulin secretion are altered with ageing. 14.7.4 Hypoglycaemia and hypokalaemia Diabetes 1985; 34 (3): 275–80. These complications are less common with current 10. Schramm VA, Push HJ Franke H and Haubitz I. Hor- low-dose insulin therapy [29, 61, 62]. The potassium monal adaptive capacity in old age: behaviour of hor- deficit should be adequately corrected and 5% dextrose monal parameters after insulin hypoglycaemia in young added to infusion fluids as soon as the plasma glucose and old patients. Fortsch. Med. 1981; 99: 1255–60. level falls below 12–14 mmol l−1. 11. Kalk WJ, Virik AI, Pimstone BL, Jackson WPU, Marker JC, Cryer PE and Clutter WE. Growth hormone response to insulin hypoglycaemia in the elderly. J Gerontol 1973; 28: 431–3. 14.8 Conclusions 12. Marker JC, Cryer PE and Clutter WE. Attenuated glucose recovery from hypoglycaemia in the elderly. Today, much remains to be done to reduce the inci- Diabetes 1992; 41: 671–8. dence of DKA and HHS, and to improve the outcome 13. Lenters KM, Ortiz RJ, Herman WH, Zobel D and of patients with these conditions. Although it has been Halter JB. (1990) Impaired glucose counterregulation suggested that the mortality rate associated with these in response to insulin-induced hypoglycaemia in the complications is decreasing, it is still considered ex- elderly. Clin Res. 38: 27OA. cessive [73]. The various factors that can precipitate 14. Meneilly GS, Cheung E and Tuokko H. Altered re- hyperglycaemic decompensation in patients with dia- sponses to hypoglycaemia of healthy elderly people. J betes should alert the physician to an early diagnosis Clin Endocrinol Metab 1994; 78: 1341–8. and prompt therapy. 15. Mitraku A, Ryan C, Veneman T, Mokan M, Jenssen T, Kiss I, Durrant J, Cryer P and Gerich J. Hierarchy of glycemic thresholds for counterregulatory hormone References secretion, symptoms and cerebral dysfunction. Am J Physiol 1991; 266: E67–74. 1. The Diabetes Control and Complications Trial Re- 16. Matyka K, Evans M, Lomas J, Cranston I, Macdonald search Group. The effect of intensive treatment of dia- I and Amiel SA. Altered hierarchy of protective re- betes on the development and progression of long-term sponses against severe hypoglycemia in normal aging complications in insulin-dependent diabetes mellitus. N in healthy men. Diabetes Care 1997; 20 (2): 135–41. Engl J Med 1993; 329: 977–86. 17. Shorr RI, Ray WA, Daugherty JR and Griffin MR. 2. UK Prospective Diabetes Study (UKPDS) Group. In- Incidence and risk factors for serious hypoglycemia tensive blood-glucose control with sulphonylureas or in older persons using insulin or sulfonylureas. Arch insulin compared with conventional treatment and risk Intern Med 1997; 157: 1681–6. of complications in patients with type 2 diabetes 18. Saudek CD and Golden SH. Feasibility and outcomes (UKPDS 33). Lancet 1998; 352: 837–53. of insulin therapy in elderly patients with diabetes 3. Teo SK and Ee CH. Hypoglycaemia in the elderly. mellitus. Drugs Aging 1999; 14: 375–85. Singapore Med J 1997; 38: 432–4. 19. Madsbad S. Insulin analogues: have they changed in- 4. Lassmann-Vague V. Hypoglycaemia in elderly diabetic sulin treatment and improved glycaemic control? Dia- patients. Diabetes Metab 2005; 31: 5S51–5S55. betes Metab Res Rev 2002; 18 (1): S21–8. 206 CH 14 METABOLIC DECOMPENSATION IN THE ELDERLY

20. Hermansen K, Colombo M, Storgaard H, O Stergaard 33. Glaser N. Pediatric diabetic ketoacidosis and hyper- A, Kolendorf K and Madsbad S. Improved postprandial glycemic hyperosmolar state. Pediatr Clin North Am. glycemic control with biphasic insulin aspart relative to 2005; 52 (6): 1611–35. biphasic insulin lispro and biphasic human insulin in 34. Alberti KGMM. (2001) Diabetic acidosis, hyperosmo- patients with type 2 diabetes. Diabetes Care 2002; 25: lar coma, and lactic acidosis. In KL Becker (ed.). Prin- 883–8. ciples and Practice of Endocrinology and Metabolism. 21. Garber AJ, Clauson P, Pedersen CB and Kølendorf 3rd ed. Lippincott Williams & Wilkins, Philadelphia, K. Lower risk of hypoglycemia with insulin detemir pp. 1438–50. than with neutral protamine hagedorn insulin in older 35. Kitabchi AE and Murphy MB. (2002) Hyperglycemic persons with type 2 diabetes: a pooled analysis of Phase crises in adult patients with diabetes mellitus. In: Ox- III trials. J Am Geriatr Soc 2007; 55 (11): 1735–40. ford Textbook of Endocrinology. JA Wass, SM Shalet 22. Motta M, Bennati E, Ferlito L, Passamonte M, Cardillo and SA Amiel (eds). Oxford University Press, Oxford, E and Malaguarnera M. A review on the actual trends pp. 1734–47. of insulin treatment in elderly with diabetes. Arch 36. Meneilly GS and Tessier D. Diabetes in the elderly. Gerontol Geriatr. 2008, 47 (1), 151–61. Diabet Med 1995; 12: 949–60. 23. Moses R. A review of clinical experience with prandial 37. MacIsaac RJ, Lee LY, McNeil KJ, Tsalamandris C glucose regulator, repaglinde in the treatment of type 2 and Jerums G. Inﬂuence of age on the presentation diabetes. Exp Opin Pharmacother 2001; 67: 1455–67. and outcome of acidotic and hyperosmolar diabetic 24. Lecomte P. Diabetes in the elderly: consideration emergencies. Intern Med J. 2002; 32 (8): 379–85. for clinical practice. Diabetes Metab 2005; 31: 38. Greene DA. Acute and chronic complications of dia- 5S51–5S55. betes mellitus in older patients. Am J Med 1986; 80: 25. Kitabchi AE, Umpierrez GE, Murphy MB, Barrett EJ, 39–53. 39. Singh I and Marshall MC Jr. Diabetes mellitus in the Kreisberg, RA Malone JI and Wall BM. Management elderly. Endocrin Metab Clin N Am 1995; 24: 255–72. of hyperglycemic crises in patients with diabetes. Dia- 40. Foster DW and McGarry JD. The metabolic derange- betes Care 2001; 24: 131–53. ments and treatment of diabetic ketoacidosis. N Engl J 26. Ennis ED, Stahl E and Kreisberg RA. The hyperosmo- Med 1983; 309: 159–69. lar hyperglycemic syndrome. Diabetes Rev 1994; 2: 41. Ennis ED and Kreisberg RA. (2000) Diabetic ketoaci- 115–26. dosis and the hyperglycemic hyperosmolar syndrome. 27. Kitabchi AE, Fisher JN, Murphy MB, et al. (1993) In: D. LeRoith, SI Taylor and JM Olefsky (eds). Dia- Diabetic ketoacidosis and hyperglycemic, hyperosmo- betes mellitus. A fundamental and clinical text.: Lip- lar nonketotic state. In: CR Kahn and GC Weir (eds). pincott Williams & Wilkins, Philadelphia, pp. 336–47. Joslin’s Diabetes Mellitus Textbook. Lea & Febiger, 42. Halperin ML, Marsden PA, Singer GG and West ML. Philadelphia, pp. 753–60. Can marked hyperglycemia occur without ketosis? Clin 28. Fishbein H and Palumbo PJ. (1995) Acute metabolic Invest Med 1985; 8: 253–6. complications in diabetes. In: National Diabetes Data 43. Gerich JE, Martin MM and Recant L. Clinical and Group. Diabetes in America. National Institutes of metabolic characteristics of hyperosmolar non-ketotic Health, National Institute of Diabetes and Digestive and coma. Diabetes 1971: 20: 228–38. Kidney Diseases, Bethesda, MD, pp. 283–91. 44. Berger W and Keller U. Treatment of diabetic ke- 29. Kitabchi AE and Wall BM. Diabetic ketoacidosis. Med toacidosis and non-ketotic hyperosmolar diabetic coma. Clin North Am 1995; 79: 9–37. Balliere’s Clin Endocrinol Metab 1992; 6: 1–22. 30. Kitabchi AE and Nyenwe EA. Hyperglycemic crises 45. Adrogue HJ, Wilson H, Boyd AE, III, Suki WN and in diabetes mellitus: diabetic ketoacidosis and hyper- Eknoyan G. Plasma acid-base patterns in diabetic ke- glycemic hyperosmolar state. Endocrinol Metab Clin toacidosis. N Engl J Med 1982; 307: 1603–10. N Am 2006; 35: 725–51. 46. Hillman K. Fluid resuscitation in diabetic 31. Kitabchi AE, Umpierrez GE, Murphy MB and Kreis- emergencies – a reappraisal. Intensive Care Med berg RA. Hyperglycemic crises in adult patients with 1987; 13: 4–8. diabetes. A consensus statement from the American Di- 47. Kreisberg RA. (1990) Diabetic ketoacidosis. In: M abetes Association. Diabetes Care 2006; 29: 2739–48. Rifkin and D Porte (eds). Diabetes mellitus: theory and 32. Kitabchi AE and Fisher JN. (1981) Insulin therapy practice. Elsevier Science, New York, pp. 591–603. of diabetic ketoacidosis: Physiologic versus pharmaco- 48. Umpierrez GE, Khajavi M and Kitabchi AE. Review: logic doses of insulin and their routes of administration. diabetic ketoacidosis and hyperglycemic hyperosmo- In Handbook of Diabetes Mellitus, M. Brownlee (ed.). lar nonketotic syndrome. Am J Med Sci 1996; 311: Vol. 5. Garland ATPM Press, New York, pp. 95–149. 225–33. 14.8 CONCLUSIONS 207

49. Gonzalez-Campoy JM and Robertson RP. Diabetic ke- 61. Fleckman AM. Diabetic ketoacidosis. Endocrinol toacidosis and hyperosmolar nonketotic state: gain- Metab Clin North Am 1993; 22: 181–207. ing control over extreme hyperglycemic complications. 62. Kitabchi AE. Low-dose insulin therapy in diabetic ke- Postgrad Med 1996; 99: 143–52. toacidosis: Fact or ﬁction? Diabetes Metab Rev 1989; 50. Braaten JT. Hyperosmolar nonketotic diabetic coma: 5: 337–63. diagnosis and management. Geriatrics 1987; 42: 63. Burghen GA, Etteldorf JN, Fisher JN and Kitabchi 83–92. AQ. Comparison of high-dose and low-dose insulin 51. Lorber D. Nonketotic hypertonicity in diabetes melli- by continuous intravenous infusion in the treatment of tus. Med Clin North Am 1995; 79: 39–52. diabetic ketoacidosis in children. Diabetes Care 1980; 52. Brandenburg MA and Dire DJ. Comparison of arterial 3: 15–20. and venous blood gas values in the initial emergency 64. Umpierrez GE, Latif K, Stoever J, et al. Efﬁcacy of department evaluation of patients with diabetic ketoaci- subcutaneous insulin lispro versus continuous intra- dosis. Ann Emerg Med 1998; 31: 459–65. venous regular insulin for the treatment of patients with 53. Wachtel TJ, Tetu-Mouradjian LM, Goldman DL, Ellis diabetic ketoacidosis. Am J Med. 2004; 117: 291–6. SE and O’Sullivan PS. Hyperosmolarity and acidosis 65. Umpierrez GE, Cuervo R, Karabell A, et al. Treatment in diabetes mellitus: a three-year experience in Rhode of diabetic ketoacidosis with subcutaneous insulin as- Island. J Gen Intern Med 1991; 6: 495–502. part. Diabetes Care 2004; 27: 1873–8. 54. Wiggam MI, O’Kane MJ, Harper R, Atkinson 66. Barnes HV, Cohen RD, Kitabchi AE and Murphy MB. AB, Hadden DR, Trimble ER, et al. Treatment of (1990) When is bicarbonate appropriate in treating diabetic ketoacidosis using normalization of blood metabolic acidosis including diabetic ketoacidosis? In: 3-hydroxybutyrate concentration as the endpoint of G Gitnick, HV Barnes, TP Duffy et al. (eds). Debates emergency management. Diabetes Care 1997; 20: in Medicine. Yearbook, Chicago, p. 172. 1347–52. 67. Okuda Y, Adrogue HJ, Field JB, Nohara H and Ya- 55. Tallis RC and Fillit HM. (2003) Geriatric Medicine mashita K. Counterproductive effects of sodium bi- and Gerontology, 6th edition. Churchill Livingstone, carbonate in diabetic ketoacidosis. J Clin Endocrinol London. Metab 1996; 81: 314–20. 56. American Diabetes Association. Hyperglycemic crises 68. Oh MS, Carroll HJ and Uribarri J. Mechanism of nor- in patients with diabetes mellitus. Diabetes Care 2003; mochloremic and hyperchloremic acidosis in diabetic 26 (Suppl. 1): S109–17. ketoacidosis. Nephron 1990; 54: 1–6. 57. Lee M, Gardin JM, Lynch JC, et al. Diabetes mellitus 69. Adrogue HJ, Eknoyan G and Suki WK. Diabetic ke- and echocardiographic ventricular function in elderly toacidosis: role of the kidney in the acid-base home- men and women. The Cardiovascular Health Study. ostasis re-evaluated. Kidney Int 1984; 25: 591–8. Am Heart J 1997; 133: 36–43. 70. Madias NE, Homer SM, Johns CA and Cohen 58. Samos LF and Roos BA. Diabetes mellitus in older JJ. Hypochloremia as a consequence of anion gap persons. Med Clin N Am 1998; 82 (4): 791–803. metabolic acidosis. J Lab Clin Med 1984; 104: 15–23. 59. Marshall SM, Walker M and Alberti KGM. (1997) 71. Oh MS, Banerji MA and Carroll HJ. The mechanism Diabetic ketoacidosis and hyperglycemic non-ketotic of hyperchloremic acidosis during the recovery phase coma. In: KGM Alberti, P Zimmet and RA DeFronzo of diabetic ketoacidosis. Diabetes 1981; 30: 310–13. (eds). International Textbook of Diabetes Mellitus. 72. Paton RC. Haemostatic changes in diabetic coma. Di- John Wiley & Sons, New York, pp. 1215–29. abetologia 1981; 21: 172–7. 60. Ennis ED, Stahl EJ and Kreisberg RA. (1997) Dia- 73. American Diabetes Association (1996) Acute compli- betic ketoacidosis. In: D Porte Jr and RS Sherwin (eds). cations. In: Diabetes 1996 Vital Statistics. American Diabetes Mellitus: Theory and Practice. Elsevier, Am- Diabetes Association, Alexandria, VA, pp. 29–44. sterdam, pp. 827–44. 15 Nutritional Perspectives: Diabetes in Older People

Begona˜ Molina1 andAlanJ.Sinclair2 1Servicio de Endocrinolog´ıa y Nutrici´on, Hospital Infanta Cristina, Avda 9 de Junio, Parla, Madrid, Spain 2Bedfordshire and Hertfordshire Postgraduate Medical School, Putteridge Bury Campus, Luton, UK

impaired fasting glucose (IFG) is also considered, the Key messages prevalence of both conditions in this population rises to 33.6%. Commencing at age 60, the lifetime risk of • Although many older people with diabetes subsequently developing diabetes is also high – 22.4% are obese, a proportion is recognized to have for women and 18.9% for men [1]. malnutrition or lesser degrees of nutritional To facilitate successful ageing in patients with dia- impairment. betes, it is important to recognize the heterogeneity in • The use of oral supplements to increase energy the spectrum of health among older patients, so that an and nutrient intake should be considered at an appropriate focusing of health interventions can take early stage in nutritional planning, and may lessen several serious adverse outcomes. place. This must coincide with an assessment of func- • Nutritional recommendations for older peo- tional status and the presence of geriatric syndromes. ple should particularly take into consideration As a person’s life expectancy may be shorter than the food preferences, cultural background, socioe- time needed to beneﬁt from an intervention, it is im- conomic factors and the support structure that is portant to prioritize treatment strategies. A realistic and available to underpin management. individualized approach is mandatory, as any one of • Nutritional interventions in older people must the comorbidities may take precedence over the actual become routine following a proper assessment; management of the diabetes [2]. the presence of other comorbidities and chronic Several physiological changes associated with age- disease make this of paramount importance. ing predispose to the development of diabetes; moreover, once diabetes has been diagnosed, insulin resistance is invariably present and may be associated with obesity [3]. In contrast, it has also been shown that 15.1 Introduction there is a high prevalence of malnutrition in the elderly, especially with regards to long-term care and In adults aged over 60 years, diabetes mellitus (DM) hospital settings [4]. The impact of malnutrition could has reached almost epidemic proportions, affecting 8.6 be more deleterious in elderly patients with diabetes if million (18.3% of the population in the USA). When nutritional counselling is poor.

Diabetes in Old Age Third Edition Edited by Alan J. Sinclair © 2009 John Wiley & Sons Ltd. ISBN: 978-0-470-06562-4 210 CH 15 NUTRITIONAL PERSPECTIVES: DIABETES IN OLDER PEOPLE 15.2 Basis of nutritional support Changes in body weight with increasing age The results of large cross-sectional studies have shown According to Peake [5], the following are the basic that both body weight and body mass index (BMI) questions to bear in mind when designing nutrition pro- increase throughout adult life until the age of about grammes for patients with hyperglycaemia or diabetes: 50–60 years, after which they decline. A substan- • What is the nutritional status of the patient? tial minority of older people undergoes a marked weight change over time. In one study [7], 17% of • Is artificial nutritional support indicated? home-dwelling people in the USA aged >65 years ≥ • Can the patient’s requirements be met by oral nutri- lost 5% of their initial bodyweight over 3 years, tional support? whereas 13% gained 5% or more. A BMI range of 24–29 kg m−2 has been suggested as appropriate for • Are there any disorders of gastrointestinal motility? the elderly population [8], especially for individuals • Should enteral or parenteral feeding be used? aged over 70 years, as this higher than conventional BMI range has been associated with lower mortality • What will be the best formula? rates. Morley suggested that weight reduction be con- • What is an acceptable blood glucose range? sidered only in those patients who were 20% above their desirable body weight, at which time a BMI of • What insulin regimen will be most effective? 29 kg m−2 may be a safer target for older people to achieve [9]. A weight loss of 5–10% from initial body 15.2.1 Normal ageing and nutritional weight is known to benefit blood pressure, glycaemic status control and lipid profiles. Changes in appetite and food intake with increasing age Changes in body composition with increasing age A recognized correlation exists between an impaired With ageing, there is a progressive increase in fat and sense of smell and a reduced interest in (and intake of) decrease in fat-free mass, which is due to a loss of food. Hence, as the senses of taste and smell deteriorate skeletal muscle; such loss may be up to 3 kg of lean with age – by 60–80% in some cases – people will body mass per decade after the age of 50 years. Con- tend to become less hungry and eat less as they grow sequently, at any given weight, older people not only older. Healthy older persons are less hungry, consume have more body fat than do young adults but it is smaller meals more slowly, eat fewer snacks between also located in different places. The increase in body meals, and become satiated more rapidly after eating a fat with ageing is multifactorial in origin; a decreased standard meal than do younger persons [6]. However, physical activity is a major cause, with contributions as the decrease in energy intake is greater than the from reduced growth hormone secretion, declining sex decrease in energy expenditure, there is a tendency for hormone action, and reduced resting metabolic rate and the elderly to lose body weight. This physiological, thermic effects of food. In older people a greater pro- age-related reduction in appetite and energy intake has portion of body fat is intrahepatic, intramuscular and been termed ‘the anorexia of aging’. intra-abdominal (versus subcutaneous); these changes Contributing to this are age-related changes with are associated with increased insulin resistance and, regards to gut motility (oesophagus, stomach and large therefore, are likely to be associated with adverse intestine), while gastric emptying is slowed down in metabolic outcomes [10]. aged persons, but without causing the problems of The anorexia–sarcopenia–sachexia triad neuropathy as seen in diabetic subjects. Together with in the elderly a diminished distensibility of the stomach, this may induce an earlier satiety and therefore a minimization Malnutrition in the elderly seems to be caused by a di- of food intake. Compared to the many other reasons for minished food intake leading to anorexia and, through malnutrition in the elderly, these changes play only a inflammatory processes, inducing cachexia. Although minor role; however, in elderly patients with diabetes, anorexia, sarcopenia and cachexia partly have com- gastroparesia might represent an important issue in mon backgrounds, anorexia mainly means a deficient nutritional terms. caloric intake first leading to a loss of body fat mass. 15.2 BASIS OF NUTRITIONAL SUPPORT 211 This is distinct from sarcopenia, where primarily lean from a metabolic point of view, but require different body mass (muscles) are lost, leading to loss of func- nutritional approaches. tionality (Table 15.1). Cachexia, on the other hand, leads to both a loss of fat and muscle mass, as seen 15.2.2 Undernutrition in older people in patients where cancer is at an advanced stage [11]. The resultant weight loss and associated undernutrition Prevalence can contribute further to adverse outcomes, because a Protein-energy malnutrition is common in the el- loss of body weight beyond the age of 60 years is derly. Studies conducted in developed countries have disproportionately of lean body tissue, predominantly shown that up to 15% of community-dwelling and skeletal muscle. When excessive, this may lead to sar- home-bound elderly, between 23% and 62% of hos- copenia (defined as muscle mass more than two stan- pitalized patients, and up to 85% of nursing homes dard deviations below the sex-specific young-normal residents suffer from this condition [6]. Protein-energy mean), which is present in 6–15% of people aged malnutrition is associated with impaired muscle func- >65 years [12]. Unlike the loss of fat tissue, a loss tion, a decreased bone mass, immune dysfunction, of skeletal muscle is associated with metabolic, phys- anaemia, reduced cognitive function, poor wound heal- iological and functional impairments; with disability, ing, delayed recovery from surgery and, ultimately, including increased falls; and with an increased risk increased morbidity and mortality. Epidemiological of protein-energy malnutrition. In the National Health studies have shown that protein-energy malnutrition is And Nutrition Examination Survey (NHANES) III a strong independent predictor of mortality in elderly study, older people who had marked sarcopenia were people, regardless of whether they live in the commu- between 3.3-fold (women) and 4.7-fold (men) more nity or in a nursing home, are patients in a hospital, likely to have a physical disability than were those or have been discharged from hospital during the past one to two years [13–15]. This mortality is even in- with low-risk skeletal muscle mass [12]. creased in the presence of other medical diseases (e.g. Ageing per se seems to be partly triggered by in- renal failure, cardiac failure, cerebrovascular disease), flammatory processes, increasing the oxidative stress, some of which are also increased in diabetes. Two of and accumulating over the whole life-span. Malnutri- the most common markers of undernutrition and risk tion in the elderly can be present in lean and obese for morbidity and mortality are low body weight and patients. Both situations are caused by an inflamma- loss of weight. tory process described before. The first ones are mal- Estimation of the prevalence of malnutrition in el- nourished because a diminished food intake leading to derly subjects depends on the tools used to evaluate anorexia, and posterior cachexia. The second ones are the nutritional status, and on the setting of the stud- malnourished because obesity can be associated to an ied population. In the past, methods used have in- increased fat body mass, and high BMI, but to a low cluded anthropometry, reports of recent weight loss, lean body mass and sarcopenia. This lean mass has a biochemical markers, and the Mini Nutritional Assess- very low metabolic rate and a poor response to low ment or other composite nutritional evaluation tools. calorie diets that can lead to higher malnutrition. So Various studies have reported an association between both, malnourished elderly patients with diabetes and mortality and nutritional status, as assessed by BMI, obese elderly patients with diabetes, reflect two types weight loss, plasma levels of albumin or food in- of inflammatory status and are completely different take. However – and especially in older people with

Table 15.1 Pathophysiology of malnutrition. Triad of malnutrition Anorexia/ Deficient caloric intake (loss of fat mass) undernutrition Sarcopenia Protein-deficient diet and lack of physical exercise (loss of muscle mass) Cachexia Catabolic state. Inflammatory parameters increased (TNF-α, IL-1, IL-6) (loss of both fat and muscle stores) 212 CH 15 NUTRITIONAL PERSPECTIVES: DIABETES IN OLDER PEOPLE diabetes – it is important to take into account other po- mortality and body weight is a J-shaped curve, with tential predictors of adverse outcomes, such as illness increased mortality at low and high BMI-values. For severity, comorbidity and functional status. young adults, the BMIs associated with the greatest life In hospital, both age and albumin have significant expectancy are in the range of 20–25 kg m−2.TheBMI effects on mortality. In one study of hospitalized pa- that is associated with maximum life expectancy in- tients, the lower the albumin levels on admission and creases with age. The lower end of the range increases the older the patient, the higher was the risk of death to about 22–23 kg m−2, while the upper end increases [16]. With regards to food intake, a total of 102 patients to 27–28 kg m−2 for people aged >65 years. Below (21%) among a total of 497 aged ≥65 years had an av- a BMI of 22–23 kg m−2 there is a steady increase in erage daily in-hospital nutrient intake of less than 50% risk of death, mainly at BMI values <18.5kgm−2 in of their calculated maintenance energy requirements, women and 20.5 kg m−2 in men [18]. while the low nutrient intake group had a higher rate Mortality was studied as a function of BMI in of in-hospital mortality (RR = 8.0; 95% CI 2.8–22.6) 8428 hospitalized patients. In those aged 20–40 years, and 90-day mortality (RR = 2.9; 95% CI 1.4–6.1) [17]. mortality doubled in the most underweight (BMI The question is then, why should protein-energy mal- <18 kg m−2) compared to groups with BMIs between nutrition be important in old people with diabetes? It 20and40kgm−2; in patients aged 70–79 years there is not unusual to prescribe a low-energy, low-fat diet was a tripling in mortality for BMI <18 kg m−2 com- for patients with diabetes in a hospital setting, but re- pared to that in groups with BMIs between 32 and strictive diets can induce a marked reduction in the 40 kg m−2 [19] (Figure 15.1). The conclusion is that average daily nutrient intake of maintenance energy being very underweight becomes increasingly lethal as requirements, leading to a worse nutritional status. In patients age. older diabetic subjects, weight loss and malnutrition have also been associated with other adverse outcomes, Causes of undernutrition in older people such as length of stay in the hospital, hospital discharge location or time to readmission, infections, gait disor- Healthy aging is associated with a decline in energy ders, falls and fractures and poor wound healing. (food) intake, the physiological ‘anorexia of aging’, and a reduction in function of homeostatic mechanisms that function in younger people to restore food intake Low body weight in older people in response to anorectic insults. Roberts and colleagues Body weight tends to decrease after the age of 60 [20] proved this fact by underfeeding young and old years, and a loss of ≥5% of body weight is not un- men by approximately 750 kcal per day for 21 days, common in older people. The relationship between during which time both groups of men lost weight.

AGES 20 20–49 18 50–59 16 60–69 70–79 14 80–99 12 10 8 6

Predicted mortality (%) 4 2 0 10 15 20 25 30 35 40 45 50 55 Body mass index (kg m−2)

Figure 15.1 Association between BMI and mortality as a function of age in 8428 hospitalized patients. Reproduced from Ref. [19] 15.2 BASIS OF NUTRITIONAL SUPPORT 213

Following the underfeeding period, when the men were aged ≥65 years had a BMI of ≥25 kg m−2 or more, allowed to eat ad libitum, the young men ate more than and the prevalence of obesity (BMI >30) in people at baseline and quickly returned to normal weight. In in the USA who were aged >70 years increased by contrast, the old men did not compensate, returned only over one-third between 1991 and 2000, from 11.4% to to their baseline intake, and did not regain the weight 15.5% [24]. The increase in the relative risk of death that they had lost. associated with being obese is not as great in older adults as it is in young adults. In an assessment of 13 Pathologic anorexia and undernutrition studies, in which non-hospitalized people aged >65 in older people years were followed for at least 3 years [25], an as- Protein-energy malnutrition is particularly likely to sociation was found between mortality and high BMI develop in the presence of other ‘pathological’ factors, in only a few cases, and then at BMIs only above − many of which become more common with increased 27–28.5 kg m 2, with little or no increase in mortality age. As most are responsive to treatment, recognition at any BMI for people aged >75 years. Where an op- is important. Older people are more likely to live timum BMI could be identified, it usually was in the − alone than are young adults, and social isolation and range of 27–30 kg m 2. The causes of increased mor- loneliness have been associated with decreased appetite tality are essentially diabetes, hypertension, sleep ap- and energy intake in the elderly. Moreover, they tend to noea, cardiovascular disease (CVD), and an increased consume substantially more food during a meal when risk with obesity for developing certain cancers, in- eating in the company of friends than when eating cluding breast, uterus, colon and prostate. Functional alone [21]. capacity and mobility are reduced significantly in the Depression is a common problem in older people, it obese elderly, who have a lower quality of life, greater being present in 2–10% of community-dwelling older limitations of physical function, and are more likely to people and in a much greater proportion of those in be homebound [26]. Obesity is predictive of a greater institutions. Depression is more likely to manifest as rate of future disability, declines in functional status, a reduced appetite and weight loss in the elderly, and an increased admission rate to nursing homes. and is an important cause of weight loss and under- The most characteristic abnormality in glucose nutrition in this group. Depression is the cause in metabolism seen in ageing is a progressive rise in 30–36% of medical outpatients and nursing home res- fasting and postprandial plasma insulin concentration, idents who lose weight. Undernutrition – particularly which suggests that ageing is an insulin-resistant if it produces folate deficiency – may worsen depres- state. The liver maintains normal glucose levels sion. The treatment of depression is effective in pro- postprandially and during fasting; with ageing, more ducing weight gain and improving other nutritional insulin is required to appropriately regulate hepatic indices [22]. glucose production and to avoid hyperglycaemia [27]. Another relevant factor is that poor dentition and Failure of the beta cell to adequately secrete insulin ill-fitting dentures, both common in the elderly, may in the face of hepatic and peripheral (muscle and fat) limit the type and quantity of food eaten. Complaints insulin resistance leads to the development of diabetes of problems with chewing, biting and swallowing are mellitus. Moreover, hyperglycaemia itself causes a usual among nursing home residents, and those with further deterioration in muscle and liver sensitivity dentures are more likely to have a poor protein intake to insulin, and a continued impairment in beta-cell [23]. The elderly often also take multiple medications, function (a phenomenon known as ‘glucose toxicity’), which increases the risk for drug interactions that can which leads to a worsening of glucose tolerance in cause anorexia. the patient with diabetes. Insulin resistance and hyperinsulinaemia are associ- 15.2.3 Overnutrition and obesity ated with increases in total and visceral fat mass, which in older people are typical of ageing. Insulin resistance is considered to be an independent risk factor for the development Prevalence of coronary artery disease (CAD), and has been asso- A substantial number of older people in westernized ciated with hypertension, dyslipidaemia and dysfibri- countries are overweight by standard BMI criteria, and nolysis (the insulin resistance syndrome, or syndrome the trend is increasing. In 2000, 58% of Americans X), which is considered to be a major risk factor for 214 CH 15 NUTRITIONAL PERSPECTIVES: DIABETES IN OLDER PEOPLE atherosclerotic CVD, cancer and all-cause mortality, the condition being treated. Patients with hypergly- and is therefore related to a decreased life expectancy. caemia form a heterogeneous group. It is vital to A strong correlation between obesity and insulin resis- identify the aetiology of the hyperglycaemia in order tance has been demonstrated in a variety of epidemio- to tailor appropriate clinical care with clear medical logical studies; the fact that insulin sensitivity improves and nutritional management objectives. The stress re- with weight loss in obese patients suggests a causal- sponse to trauma/illness commonly exacerbates hyper- ity relationship between obesity and insulin resistance. glycaemia in patients with pre-existing diabetes, and It is speculated that the changes in body composition not infrequently produces significant hyperglycaemia in the process of ageing are responsible for the devel- in previously euglycaemic patients due to the increase opment of insulin resistance, and consequently for the in circulating counter-regulatory hormones (glucagon, increased prevalence of type 2 diabetes with ageing. cortisol, growth hormone, adrenaline), resulting in an All of these physiopathological considerations increase in hepatic glucose production, a decrease in should be borne in mind when treating the peripheral glucose uptake and an increase in insulin insulin-resistant elderly patient with diabetes. Most resistance. However, when assessing a patient with di- therapeutic strategies are designed to induce an abetes for nutritional support, their recent glycaemic improvement in carbohydrate homeostasis by weight control and treatment are highly relevant, as poor gly- reduction (decreased fat), directly improving insulin caemic control can compromise the patient’s overall sensitivity, and finally decreasing the deleterious nutritional status. effects of glucose toxicity. A detailed discussion of methods that are used to diagnose undernutrition in older people with diabetes Should overweight older people be advised is beyond the scope of this chapter; there is no ‘gold to lose weight? standard’, and multiple methods have been used. The most important thing is to be aware of the diagnosis The adverse effects of obesity, reduced life expectancy and to use effective tools for screening. An assessment at high BMIs up to the age of at least 70 years, and im- is a more detailed and longer examination carried provements in function that are associated with weight out by specially trained staff (clinician, dietician, or loss, must each be balanced against the detrimental ef- nutrition nurse) in those patients screened as being fects of weight loss on muscle mass and bone density, at nutritional risk, according to the ESPEN guidelines and the proven association between all-cause weight [29]. In addition, an intervention plan should be based loss and increased mortality in older people, even on the results of the screening tool. those who are overweight initially. It could be that All elderly people aged >65 years should be an older person overweight to a degree that reduces screened once a year and weighed at regular life expectancy might be likely to die sooner if at- intervals, particularly those in nursing homes or other tempting to lose weight and succeeding than remaining institutions, because a weight loss of more than 5% is weight-stable. Studies involving younger adults have usually a key indicator. A BMI <22 kg m−2 suggests found that intentional weight loss can reduce mortal- undernutrition, which is particularly likely if the BMI ity in those who have obesity-related health problems is <18.5kgm−2. (type 2 diabetes, ischaemic heart disease, hypertension) Reduced serum albumin, haematocrit, lymphocyte [28]. Available evidence suggests that it is safe to rec- count and serum folate are among the factors found ommend weight loss to overweight older people who to be associated with the risk for undernutrition and have obesity-related morbidities, particularly reduced poor outcome [30]. Among the most widely used mobility and function. Indeed, this group seems to have outpatient screening tools for undernutrition risk are the most to gain. the Mini Nutritional Assessment [31]. The short form of the Mini Nutritional Assessment (MNA) fulfils some of these expectations, having been derived from 15.3 Nutritional assessment the original version of the MNA by identifying six items that are strongly correlated with the conventional Before nutritional support is initiated, the indications, nutritional assessment of experienced physicians. If the aims/objectives and ideal route of nutritional support score amounts to 11 or less, the patient is classified must be determined, based on both the nutritional as- as at risk of malnutrition and the full MNA must sessment of the patient and the clinical features of be carried out. In a multitude of studies, the MNA 15.3 NUTRITIONAL ASSESSMENT 215 has been shown to correlates well with nutritional nutritional screening of all hospital patients without intake, anthropometry, laboratory data, functionality, age restriction, and its application seems to be most morbidity, length of stay and mortality [32]. In nursing appropriate in this setting [33, 34]. If the total score homes and hospitals there will be more obstacles that is ≥3, then some form of nutritional support should may hinder the successful application of the MNA. be started. In Table 15.2 are shown the main validated Many of these patients are not capable of cooperating, methods for nutritional screening. and under these circumstances the Nutritional Risk For special subgroups within this population who Screening (NRS) may be used as an alternative. In show a high prevalence of malnutrition, it is advised comparison with the MNA, the NRS emphasizes more to proceed directly with the assessment tool. Examples the severity of concomitant diseases in the screening might be frail elderly living largely independently in for risk. The NRS has been recommended for the the community but relying on social services, elderly

Table 15.2 Validated methods for nutritional screening. Technique Use Parameters studied Mini Nutritional Evaluation of nutritional status, validated Anthropometrics: weight, height, arm and Assessment (MNA) in the elderly; for use in outpatient calf circumferences and weight loss. [31] setting, community and nursing homes General assessment: mobility, lifestyle and if possible. medication. Dietary assessment: food and fluid intake, number of meals, and autonomy of feeding. Subjective assessment: the patient’s perception of their own health and nutrition. Nutrition Risk Index Evaluation of nutritional status in the A 16-item questionnaire including: (NRI) [33] elderly for all hospital patients without mechanics of food intake, dietary age restriction. restrictions, morbid conditions affecting food intake, discomfort associated with outcome of food intake and significant changes in dietary habits. Malnutrition Universal Identifies adults, who are malnourished, at A five-step screening tool using: height and Screening Tool risk of malnutrition, or obese; for use weight, percentage unplanned weight loss, (MUST) [37] in hospitals, community and other care acute disease effect together to obtain settings. overall risk of malnutrition. Subjective Global Identifies risk of developing Nutritional history: previous 6 months’ Assessment (SGA) [38] nutrition-related complications. weight loss, pattern of dietary intake, presence of gastrointestinal symptoms, functional capacity. Physical examination: loss of subcutaneous fat, muscle wasting, loss of fluid. Prognostic Nutritional Identifies increased risk of post-surgical Combines anthropometry, Index (PNI) [39] complications delayed-hypersensitivity skin test and plasma protein levels – expressed as a single value.

Adapted from [36] Reilly, H.M. (1996) Screening for nutritional risk. Proceedings of the Nutrition Society, 55, 841–53. [37] Malnutrition Advisory Group (MAG). MAG guidelines for detection and management of malnutrition. 2000, Redditch, UK: British association for Parenteral and Enteral Nutrition. Available at: http://www.bapen.org.uk/must tool.html [38] Detsky, A.S., McLaughlin, J.R., Baker, J.P., Johnston, N., Whittake, S., Mendlson, R.A. and Jeejeebhoy, K.N. (1987) What is subjective global assessment of nutritional status? Journal of Parenteral and Enteral Nutrition, 11, 8–13. [39] Buzby, G.P., et al. (1980) Prognostic nutritional index in gastrointestinal surgery. American Journal of Surgery, 139, 160–7. 216 CH 15 NUTRITIONAL PERSPECTIVES: DIABETES IN OLDER PEOPLE

with extensive comorbidities as diabetes, nursing home Table 15.3 Mnemonic ‘Meals on Wheels’. Adapted residents, and elderly hospital patients [29]. from Ref. [35]. Initially, it is essential to obtain an accurate medical • Medications (e.g. digoxin, theophylline, fluoxetine) history of the patient. The most important clinical as- • pect leading to the diagnosis of malnutrition certainly Emotional causes (depression) addresses the course of the patient’s weight and here • Alcoholism especially weight loss, which should be expressed ei- • Late-life paranoia ther in kilograms or as a percentage of the patient’s • Swallowing problems (dysphagia) usual weight. The interval since the weight loss started • should also be explored. The patient should be asked Oral problems whether they have suffered a loss of appetite, and/or • Nosocomial infections (TB, Clostridium difficile, He- it may be helpful to question the patient about re- licobacter pylori) strictive diets and the consumption of alcohol or to- • Wandering (dementia) bacco. Extra attention should be paid to physical signs • Hyperthyroidism, hyperparathyroidism, hypoadrenal- of over-malnutrition (muscle atrophy, loss of subcuta- ism neous fat and peripheral oedema as a consequence of • hypoprotidaemia). Enteral problems (malabsorption) Since functional status in the elderly is closely • Eating problems (inability to self-feed) correlated with their nutritional status, it is usually • Low-salt, low-fat diet advisable to determine their basic and instrumental • Shopping and social problems activities of daily living (ADLs and IADLs). Some additional information about the living conditions and the social relationships of the patient must also be obtained. should also be assessed using a combination of all the John Morley’s mnemonic MEALS ON WHEELS anthropometric data (Table 15.5). summarizes a variety of treatable pathological causes The most widely used laboratory parameter to as- [35]. Chronic diseases and medications can also be sess the nutritional status has been serum albumin, risk factors for malnutrition in the elderly through although such levels may be influenced by a wide va- their effects on appetite, food intake, gastrointestinal riety of acute and chronic inflammatory and malignant function and metabolism (Tables 15.3 and 15.4). conditions; that is, serum albumin is a marker of dis- Anthropometric measurements are an essential part ease severity rather than of malnutrition. Ageing, as of the nutritional assessment of the elderly [40], well as hepatic and renal dysfunction, can also cause a and comprise the determination of body height, body decrease in serum albumin levels. A further disadvan- weight, circumference of the upper arm and calf, and tage is that serum albumin has a long half-life of 18 measurement of the triceps skin fold. Special condi- days. Although a diminished serum albumin is rarely tions such as oedema, ascites, pleural effusion and the consequence of a poor nutritional status alone, it loss of body parts must be taken into account. Serial may serve as a marker of the severity of a disease weights over days and weeks are more valuable than which itself carries a risk of developing malnutrition. single measurements, and provide important informa- Alternative parameters are transferrin, transthyretin, tion in the short term concerning fluid balance, and in retinol-binding protein and insulin growth factor-1, but the long term concerning changes in nutritional sta- these will clearly increase laboratory costs. tus. The BMI is a poorer reflection of changes in body Laboratory examinations are not an essential com- composition in the elderly than in a younger popula- ponent of the diagnosis of malnutrition in the elderly. tion. For adults aged <65 years, the cut-off point below In a recent study, it was noted that weight loss and an- which malnutrition is highly probable has been set at thropometric data showed a stronger correlation with 18.5kgm−2, while in the elderly, for prognostic rea- life-threatening complications among geriatric hospital sons, the cut-off point is usually set at 20–22 kg m−2 patients than did either albumin or transthyretin [17]. [8]. With regards to the anthropometric parameters, an Bioelectrical impedance analysis (BIA) and individual patient should not be judged to be malnour- dual-energy X-ray analysis (DXA) measurements ished on the basis of only one pathological value, but remain interesting fields of ongoing research in body 15.3 NUTRITIONAL ASSESSMENT 217

Table 15.4 Chronic diseases and drugs that may cause malnutrition in the elderly. Chronic diseases Drugs • Chronic cardiac failure • ACE-inhibitors • Chronic pulmonary diseases • Analgesics • Cancer • Antacids • Chronic infectious diseases • Anti-arrhythmic drugs • Gastrointestinal diseases • Antibiotics • Diabetes • Anti-epileptic drugs • Severe osteoarthritis • Antidepressants • Hypothyroidism/hyperthyroidism • β-blocking agents • Cerebral ischaemia • Calcium channel-blocking agents • Intracerebral bleeding • Digoxin/digitoxin • Pressure ulcers • H2-receptor antagonists • Parkinson’s disease • Laxatives • Dementia • NSAIDs • Depression • Oral antidiabetic substances • Potassium • Corticosteroids

Table 15.5 Essential parameters for the diagnosis of these need to be individually modified with considera- malnutrition in the elderly. tion of important factors, such as functional and mental ability and polypharmacy when prescribing diet or nu- • Weight loss (expressed in kg or percentage of former/usual weight) tritional support to this age group. The ADA position statement concerning nutritional intervention for older • Oral intake (simple documentation by e.g. the eye-ball adults with diabetes focuses only on weight manage- method) ment and physical activity, plus a multivitamin sup- − • Body mass index <22 kg m 2 (showing an acceptable plement recommendation for those older adults with association with body fat stores) reduced energy intake. • Calf circumference <31 cm (showing a good correlation Current European recommendations are based on with muscle mass and functional status in the elderly) studies in younger age groups, which have then been extrapolated to the elderly, leading to a low quality of evidence for the specific effects in older age groups. composition, but neither is currently used in routine The most recent European recommendations for clinical practice. adults with diabetes emphasize energy balance and weight control, and recognize a wide variation in car- 15.3.1 Brief review of nutritional bohydrate intake as being compatible with good dia- guidelines betic control. The target is to help optimize glycaemic A number of reviews of nutritional guidelines in people control and reduce the risk of CVD and nephropathy. with diabetes (including older people) have been con- The quality of life of the individual person must be ducted, including ADA [41], IDF [42], ESPEN [66], considered when defining nutritional objectives, and ASPEN [69] and NICE [43]. health care providers must achieve a balance between Clinical advice has usually been based around the the demands of metabolic control, risk factor manage- recommendations for all adults with diabetes. Clearly, ment, patient well-being and safety. 218 CH 15 NUTRITIONAL PERSPECTIVES: DIABETES IN OLDER PEOPLE 15.3.2 Current dietary recommendations: fatty acids should not exceed 10% of the total en- applications to older people with ergy. Mono-unsaturated fatty acids (MUFAs) with diabetes cis-configuration in combination with carbohydrate should provide 60–70% of the total energy, and one The National Diet and Nutrition Survey of people portion of oily fish per week plus other plant sources aged ≥65 years [44, 45] showed that two-thirds of are recommended for n-3 fatty acids. However, dietary free-living elderly were overweight or obese. While supplements of fish oils (or their derived preparations) only 3% of men and 6% of women in the community and pharmacological doses of vitamins are not recom- were underweight, this figure rose to 17% for the el- mended. The daily cholesterol intake should not exceed derly in institutions. Undernutrition in acutely ill hospi- 200 mg. Plant sterol and stanol esters block the in- talized elderly patients has been estimated at 26% [46]. testinal absorption of dietary and biliary cholesterol. The following topics should be considered: body Among the general public, and in individuals with weight, physical activity and the specific micronutri- type 2 diabetes, a daily intake of 2 g of plant sterols ent composition of the diet, including carbohydrates, and stanols has been shown to lower both plasma and protein, alcohol, sodium, vitamins and minerals. Un- LDL-cholesterol [41, 42]. If these products are used, dernutrition is as much a concern in older patients they should displace – rather than be added to – the with diabetes as is obesity [47]. The first measure diet in order to avoid weight gain. Soft gel capsules in the treatment of malnutrition may be dietary coun- containing plant sterols are also available. selling. With previously malnourished patients a nutri- Although guidelines exist that outline the appropri- tional intervention should be started sooner. Phosphate, ate use of food sources, these often change, reflecting the lack of clear evidence-based suggestions. However, chromium and trace element deficiencies can cause hy- it is important to note that, while a decreasing fat in- perglycaemia by decreasing insulin sensitivity. take is warranted for weight loss, any oils used should consist of unsaturated fats (e.g. olive, corn, canola, 15.4 Energy intake: carbohydrates and certain varieties of safflower and sunflower). Sec- and fats ond, complex carbohydrates – long polymers of glucose found in starches, such as rice, potatoes and 15.4.1 General recommendations vegetables – should be the substitutes of simple sugars. ‘Diabetic’ foods are not recommended, and The optimal diet for diabetic patients remains un- non-nutritive sweeteners afford an alternative means known. Total dietary energy should be the same for of providing sweetened foods and drinks that are elderly patients as for their younger counterparts, with palatable to the elderly. For hypertension control, the no specific recommendation unless the person is over- DASH diet (Dietary Approaches to Stop Hyperten- weight or gaining weight, when a reduction in total sion) [48] suggested: (i) an emphasis on the use of energy intake is advised. Carbohydrate should be in an fruits, vegetables and low-fat dairy products; (ii) the acceptable range of 45–60% of the total energy [41]. A inclusion of whole grains, poultry and nuts; and (iii) a combined intake of 60–70% of total energy is recom- reduction in the intake of fats, red meat, sweets and mended for carbohydrate and monounsaturated with sugar-containing beverages. cis-configuration fatty acids. The source of carbohydrate does not appear to affect glycaemic control if 15.4.2 Malnourished diabetic patients used with mixed meals, but carbohydrates rich in fi- If a patient is eating very little, then the provision bre or having a low glycaemic index are particularly of palatable sugar-containing food may help to stim- recommended, while sucrose intake should not exceed ulate the appetite. The inclusion of high-fat food is 10% of the total energy. In addition, an increase in the valuable in helping the patient maximize their energy proportion of carbohydrates from fruits, vegetables and intake in smaller portions. Conversely, the inclusion legumes should be recommended. of high-fibre diets (complex carbohydrates) may limit With regards to fats, an acceptable range 25–35% food intake by causing early satiety. The provision of the total energy should be recommended, with of extra high-protein and energy snacks may be suf- saturated and trans-unsaturated fatty acids provid- ficient to meet a patient’s nutritional requirement [49]. ing <7% of the total energy [41]. Polyunsaturated These dietary modifications are likely to increase the 15.4 ENERGY INTAKE: CARBOHYDRATES AND FATS 219 intake of simple sugar and the glycaemic index of the obesity, functional limitations, duration of diabetes, or diet, but these potentially adverse changes must be the presence of other comorbid conditions [52]. Phys- offset against the risks of malnutrition. Some foods ical training also increases insulin sensitivity – that with a high energy content have a low glycaemic in- is, patients respond better to insulin injection or to dex, and it may be advantageous to encourage the endogenous insulin. intake of foods such as ice-cream, custard, yoghurt, The current guidelines from the Centers for Disease sponge cake and muffins to minimize the glycaemic Control and Prevention [53] recommend that a 30-min response. period of moderate activity on most days is achievable for many older people. Individuals should be encour- 15.4.3 Obese diabetic patients aged to do what they can achieve, to do it regularly, and The typical patient with insulin-resistant type 2 dia- to gradually build up in intensity and frequency. Even betes is obese, and the control of diabetes mellitus will very frail older people can manage certain activities be improved after only a slight weight loss. Insulin and, over time, should be able gradually to improve sensitivity increases when obese patients are on a neg- their strength. ative caloric balance. Thus, as the major goal in improving insulin sensitivity is weight loss, a decrease in caloric intake should be emphasized [50]. As fat 15.4.5 Protein contains more than twice as many calories as car- Protein intake should account for 10–20% of the total bohydrate and protein per gram, it should be lim- energy. The European Association for the Study of ited in the hypocaloric diets of overweight patients Diabetes [42] has recognized that recommendations with type 2 diabetes. Rather, patients should switch to diet drinks or water to save a significant num- for the dietary protein content is based on incomplete ber of calories, in addition to decreasing their intake evidence. Overall, the incidence of nephropathy in of simple sugars. Low-carbohydrate diets (which re- the elderly population has increased over the past 20 strict the total carbohydrate to <130 g per day) are not years, one possible explanation being that the improved recommended in the treatment of overweight/obesity, treatment of CHD and hypertension has resulted in as the long-term effects of these diets are unknown. more patients with type 2 diabetes living long enough Although they produce short-term weight loss, the to develop nephropathy and end-stage renal failure maintenance of weight loss is similar to that with (ESRF). low-fat diets, and the impact on the CVD risk profile In a recent review which examined protein ranges −1 is uncertain. between 0.3 and 0.8 g kg body weight per day, A trained dietician should instruct the patient with high protein intakes contributed to the development diabetes about appropriate dietary strategies. This of nephropathy [54]. Reducing protein intake appears approach, which often includes the spouse, may to slow the progression to renal failure, although the restructure the patient’s eating habits – an action level of restriction that is both effective and acceptable that is critical for the long-term successful dietary to patients is unknown. Current guidelines are based treatment of diabetes. mainly on individuals with type 1 diabetes, and often use proxy indicators such as creatinine clearance rather 15.4.4 Exercise than hard clinical end-points such as time to dialysis or The benefits of exercise for all diabetics have been death from ESRF. The current European recommenda- well documented, and are irrespective of body weight tions suggest that patients with diabetes, who exhibit or age. Exercise will lead to an improvement in evidence of microalbuminuria or established nephropa- metabolic and cardiovascular risk factors, as well as thy, should have a protein intake at the lower end of − improving strength, flexibility, balance and function the normal range (0.7–0.9 g kg 1 body weight per day). [51]. In an eight-year prospective study from the NHIS Protein intakes below this level increase the risk of of 2896 adults with diabetes, walking for at least 2 h malnutrition during chronic illness or catabolic states. each week was associated with a 39% lower all-cause In some elderly patients with diabetes, the balance be- mortality rate and a 34% reduction in coronary heart tween the risk of malnutrition and the possible benefits disease (CHD) mortality. The magnitude of these of a reduced protein intake to delay nephropathy must benefits persisted after controlling for age, gender, be carefully assessed. 220 CH 15 NUTRITIONAL PERSPECTIVES: DIABETES IN OLDER PEOPLE 15.4.6 Fibre 15.4.9 Vitamins and minerals Fibre is a particularly important component to encour- Elderly persons with diabetes are at risk of micronutri- age in an older person’s diet. Soluble fibre such as oat ent deficiency, whether due to a low food intake, to the bran, pectin and guar lowers the plasma glucose lev- presence of chronic disease, or to drugs. The second els and may improve the plasma lipid profile. Soluble evaluation of the Euronut-SENECA study population fibre does not appear to interfere with the absorption [4] was made in 1993, when subjects were aged 74–79 of minerals in elderly patients with diabetes. Consti- years (n = 1005). Among this population, 23.9% of pation is common among diabetics, and increasing the the men and 46.8% of the women had low dietary in- fibre intake can reduce laxative use and improve bowel takes for at least one of the following micronutrients: function. The ADA[41] recommendation is 14 g per calcium, iron, retinol, β-carotene, thiamine, pyridox- 1000 kcal of intake mainly from food containing whole ine or vitamin C. Vitamin D plasma levels were low grains, this means mainly insoluble fibre. However, fi- in 36% of the men and 47% of the women. A cobal- brous foods tend to have a greater satiating effect and amin deficiency was described in 23.8% of the sub- should be advised with caution in those with a de- jects. In both institutionalized and hospitalized elderly pressed appetite. An increased fibre intake, particularly persons, the prevalence of micronutrient deficiency in the oldest elders, may cause bowel impaction if the appeared to be higher, especially for thiamine, pyri- liquid intake is poor. doxine, cobalamin, folates, vitamin C, vitamin E and selenium. Among the survey population, between 10 15.4.7 Sodium intake and 40% were shown to have multiple vitamin deficiencies, while 10% were anaemic. Notably, the levels Both, taste and smell decline with age, beginning of deficiencies were higher when an individual was around the age of 60 years but becoming more marked receiving institutionalized care. above 70 years. Salt and monosodium glutamate are The micronutrient status of elderly individuals with commonly used as taste enhancers, and can improve diabetes is controversial, with limited information dietary intake in elderly people. On the other hand, available and the recommendation that any interven- sodium intake is linked with the development or ex- tion should be carried out with caution until further acerbation of hypertension, and when salt intake is re- research is completed [56]. The authors concluded that duced the blood pressure may fall. A balance between micronutrient supplementations for people with dia- using flavour enhancers to encourage dietary intake for betes should be individualized and based on clinical underweight people, while not exacerbating hyperten- findings, dietary history and laboratory results. The sion, must be made. Sodium restriction requires a salt most important of these are discussed in the following intake of <6 g per day [41, 48]. sections.

15.4.8 Alcohol Specific mineral and vitamin deficiencies Older people are more susceptible to the effects of al- Vitamin D Vitamin D deficiency causes osteomala- cohol, and are likely to develop problems at relatively cia, rickets and myopathy, and is associated with re- lower levels of consumption, due to the age-related duced bone density, impaired mobility and an increased body composition changes. Moderate intakes of alco- rate of falls and fractures. In ambulatory older people, hol appear to benefit blood pressure, glycaemic control mobility was seen to decline markedly when serum and reduce the risk of thrombosis. Alcohol can also 25-hydroxyvitamin D levels were <40 pmol l−1,al- act as an appetite stimulant, which may be beneficial. though vitamin D supplementation reduced the rate of A large intake of alcohol has been shown to increase falls in nursing home residents (even those not deficient the risk of stroke, hypertension, hypoglycaemia and in vitamin D). Treatment with vitamin D at dosages of both lactic and ketoacidosis [55]. The intake should 700–800 IU per day, with or without calcium, reduced be restricted to 1–2 units per day for women, or 2–3 the relative risk for hip and other non-vertebral frac- units per day for men., though the lower end of these tures by 23–26% compared to calcium or placebo in ranges is probably preferable. When using insulin or ambulatory or institutionalized older persons [57, 58]. sulphonylureas it is advisable that alcohol is consumed A plasma 25-hydroxyvitamin D level <40 nmol/L together with carbohydrate-containing foods. (or <16 μg/L) is widely considered to represent 15.4 ENERGY INTAKE: CARBOHYDRATES AND FATS 221 vitamin D deficiency that is in need of treatment. skin should be sufficient for vitamin D synthesis for If the more generous definition of this condition is most adults, with the face and arms being exposed for used (serum 25-hydroxyvitamin D <80 nmol/L), a 30 min each day. Further studies are required to clarify much larger proportion of the population has the the assessment of vitamin and mineral status in elderly problem. Most circulating 25-hydroxyvitamin D patients with diabetes. derives from exposure of the skin to UV-B radiation in sunlight; the remainder is obtained via the dietary Vitamin B12 (Cobalamin) Vitamin B12 deficiency is intake of foods rich in vitamin D (predominantly more common in older people than in young adults. In oily fish), supplements and vitamin D-fortified food. the Framingham study, 11.3% of elderly subjects had −1 Dietary requirements are greater in older people due a serum vitamin B12 concentration <258 pmol l ,to- to a reduced production in the skin, decreased sun gether with elevated plasma homocysteine and methyl- exposure, an age-related thinning of the skin, and malonic acid levels, compared to 5.3% of younger other skin changes. For that reason the recommended adults. In elderly people living in institutions, the dietary reference intakes are higher for older adults; prevalence of deficiency may reach 30–40%. Because in the USA, it is 10 mg (400 IU) for people 51 to 70 the signs and symptoms of vitamin B12 deficiency of- years of age. The USA and Canada have mandatory ten are subtle (macrocytic anaemia, subacute combined vitamin D fortification of milk, and Canada also degeneration of the spinal cord, neuropathies, ataxia, requires it in margarine, whereas other countries have glossitis, and possibly dementia), there should be a low variable levels of non-mandatory fortification [58]. threshold for testing older people, in particular those Vitamin D therapy is safe, inexpensive and easy to who are malnourished, those who have a neurologic or administer. The prevalence of vitamin D deficiency is neuropsychiatric presentation that is consistent with vi- so high among the institutionalized elderly that routine tamin B12 deficiency, and those who are in institutions, supplementation with doses of 800–1000 IU per day, including psychiatric hospitals [62]. without testing, is being recommended and increas- There is evidence that homocysteine damages blood ingly adopted. The most effective form of replacement vessel walls, and that there is a significant association is oral cholecalciferol, which can be given in intermit- between increased plasma homocysteine levels and an tent boluses at intervals of one to six months, in doses increased risk for CVD. The increased prevalence of not usually totalling more than 50 000 IU per month deficiency in the elderly is due mainly to an increased or as 500–2000 IU per day [59]. rate of food cobalamin malabsorption and pernicious anaemia, which account for approximately 60–70% Calcium Guidelines for daily dietary intakes of cal- and 15–20% of cases, respectively [63]. cium, developed with the aim of optimizing bone Both, vitamin B12 and folate deficiencies coexist fre- health, usually recommend higher intakes for older quently in older people. The most common predispos- adults. Few older people achieved an adequate daily ing factor is gastric atrophy, which is present in more calcium intake (1200 mg) without taking a calcium than 40% of people aged >80 years. Numerous fac- supplement; the median daily dietary intake for Amer- tors predispose to the development of gastric atrophy, ican men and women aged ≥60 years is approximately including Helicobacter pylori infection, chronic alco- 600 mg [60]. One tablet per day of a supplement con- holism, bacterial overgrowth, the long-term ingestion taining 500–700 mg of elemental calcium is usually of metformin and antacids, and gastric bypass surgery sufficient to achieve an adequate intake, but people for obesity. with a low dietary intake should take two tablets. An Clinically apparent causes should be treated when adequate vitamin D status is essential for calcium up- possible, but a reversible cause of vitamin B12 defi- take by the gut and bone formation and remodelling; ciency often is not found, and so treatment with vitamin hence, any vitamin D deficiency should be identified B12 usually is needed for life. The recommended daily and corrected in older people who take calcium. Alter- intake of vitamin B12 is 2–5 mg in older adults. A vi- −1 natively, vitamin D should be added routinely to the tamin B12 deficiency (<150 pmol l ) that is due to calcium treatment [61]. dietary inadequacy is best treated initially with intra- The risk of fractures due to falls and osteoporosis muscular (i.m.) vitamin B12, or at least with 100 mg per when diabetes is present may be further increased by day of oral vitamin B12. Food malabsorption is treated peripheral neuropathy, autonomic neuropathy, hypo- best with i.m. vitamin B12 or possibly high-dose oral glycaemic episodes and poor eyesight. Sunlight on the vitamin B12 (e.g. 500 mg per day), whereas pernicious 222 CH 15 NUTRITIONAL PERSPECTIVES: DIABETES IN OLDER PEOPLE anaemia requires lifelong i.m. therapy [62]. When fo- symptoms of deficiency include weight loss, neu- late deficiency coexists, vitamin B12 should be given ropathy and impaired glucose tolerance. The preva- with appropriate folate doses; in any case, it is reason- lence of Cr deficiency in diabetes is uncertain, but able to coadminister a multivitamin that contains folate. Cr supplementation enhances glucose tolerance in patients with diabetes; however, the significance of Folate Foods that are rich in folate include orange this finding in elderly diabetic subjects remains to be juice, dark green leafy vegetables, peanuts, strawber- tested. ries, dried beans and peas, and asparagus. The synthetic folic acid found in vitamin supplements and fortified • Copper: the clinical relevance of elevated copper foods does not require intestinal cleavage and is ab- levels in the elderly patient with diabetes is still sorbed more readily. The recommended daily intake unknown. The recommended range of copper con- of folate and folic acid is 400 mg, with an upper limit sumption has recently been extended to 1.5–3.0 mg of 1000 mg of synthetic folic acid which, in high per day. doses, may mask the features of coexistent vitamin • Iron: currently, there appears to be no evidence of B deficiency in older people. Folate deficiency causes 12 any major alteration in iron status in diabetic patients macrocytic anaemia and increased homocysteine con- who do not have renal failure or neuropathy with centrations, and is associated with increased rates of delayed absorption, but the uptake and utilization of colorectal cancer and, possibly, also cervical cancer, iron is delayed. Although the iron status in elderly cognitive impairment, depression and dementia. The diabetic patients has not been studied, the elderly prevalence of folate deficiency among older people varies from 4% to 50%, depending on the population diabetic patient is most likely not at an increased studied, and is particularly common among persons in risk of iron deficiency in the absence of other causes institutions. Most folate deficiency is due to an inade- of iron loss. quate dietary intake; an impaired use due to drug intake • Magnesium (Mg): magnesium plays an important (e.g. methotrexate, anticonvulsants, sulfasalazine) or role in glucose homeostasis, and diabetes is asso- alcohol consumption is much less common. When fo- ciated with an increased urinary losses of this ion, late deficiency is due to dietary insufficiency, attempts especially when hyperglycaemia is present, and more should be made to improve the diet by increasing the so in elderly patients. The intake recommended for intake of fruit and vegetables. Folic acid supplements healthy adult males is 420 mg/d and for women is are also indicated to ensure treatment success, and are 320 mg/d. Open and double blind studies on the essential when the cause is not dietary, but should effects of the treatments of magnesium deficiency be excluded before starting treatment with folic acid and of magnesium depletions in geriatic populations (0.5–5 mg per day) [64]. are too scarce. Magnesium supplementation (3 g per day) resulted in a significant improvement in re- Other vitamins and minerals [65] sponse to glucose and arginine in a clinical trial in • Zinc: zinc deficiency is associated with abnormalities the elderly, but more studies are required to con- in a wide variety of biological functions, including firm this. anorexia, T-cell abnormalities, wound healing, impo- • tence and, possibly, macular degeneration. There is B and A vitamins: these are used in a variety currently no definitive evidence that zinc deficiency of metabolic functions, their primary source being causes diabetes or exacerbates glucose homeostasis. green leafy vegetables (that occasionally are lacking The recommended daily allowance (RDA) for zinc in the diet of many elderly patients). For patients for men aged >50 years remains at 15 mg per day with neuropathy, a 2-month trial of vitamin B1 or and 12 mg per day for women. For patients with leg B6 has been suggested, but the efficacy of such a ulcers, impotence or poor wound healing, a 3-month regimen has not been proven. No deficiency has course of zinc supplementation (70 mg elemental been found in most studies of serum vitamin A in zinc daily) is recommended. Overall, there is a con- the aged. sensus that nutrition is important for wound healing. • Vitamin C: serum deficiency states of vitamin C are • Chromium (Cr): chromium has an important role in commonly noted in both elderly and younger pa- the regulation of glucose and lipid metabolism, and tients with diabetes. The tissues stores of vitamin C 15.6 ARTIFICIAL NUTRITION 223 are also depleted in the presence of chronic hypergly- to avoid a reduced intake at meal times. The consump- caemia, but are associated with impaired leukocyte tion of each supplement should take no longer than function and microangiopathy. The UK Department 30 min, and it should ideally be controlled and docu- of Health currently recommends a reference nutrient mented by the nursing staff (even better, on the drug daily intake of 40 mg of vitamin C, but not routine rounds, when the consumption can be supervised). For supplementation. These vitamin C requirements are most supplements, 1 ml is equal to 1 kcal, but hyper- based on the prevention of scurvy, and further re- caloric drinks (1.5–2 kcal ml−1) are available and may search into the benefits of higher intakes is required. be useful under certain circumstances. High-protein However, ADA does not recommend routine sup- supplements may be indicated in cases of severe pro- plementation with antioxidants, such as vitamins E tein depletion. A realistic goal for energy intake from and C and carotene due to a lack of evidence and nutritional supplements is 400 kcal per day, although concern relating to long-term safety, unless for older in some cases this may need to be increased to 600 adults with a reduced energy intake in whom a daily kcal per day. multivitamin supplement might be appropriate. The sugar content of the supplement is often offset by the patient’s reduced dietary carbohydrate intake, and can be invaluable in preventing hypoglycaemia 15.5 Nutritional oral supplements during periods of poor oral intake. Spicy snacks may sometimes be helpful and induce old people with im- If dietary counselling alone proves insufficient, the paired taste to eat more [49]. In particular, among those next step may be to increase the energy and pro- ambulant, demented patients in residential homes, who tein concentration of the food. According to ESPEN tend to forget meal times and wander off, a finger-food guidelines for enteral nutrition in geriatric patients, oral buffet to which patients help themselves as they wish supplementation increases both energy and nutrient in- has proved superior in terms of total intake, to fixed take [66]. Oral supplements should be given at an early mealtimes and sit-down meals. stage, when there is evidence of insufficient intake, of weight loss exceeding 5% in a 3-month period or >10% in 6 months, or when the BMI is <22 kg m−2. 15.6 Artificial nutrition Oral supplements are available in a wide range of savoury and sweet flavours. There is also a wide variety The nutritional management of hospitalized adult pa- of presentations, including powders, pre-made carton tients with diabetes covers the aetiology of hypergly- sip feeds, glucose polymers (powders and syrups) and caemia, the effects of diabetes on nutritional status, the protein powders. These products may be nutritionally metabolic consequences of stress and specific nutrient complete. A recent meta-analysis [67] showed a re- mixes. Artificial nutritional support can also exagger- duced mortality and fewer complications among under- ate the hyperglycaemic response to stress caused by nourished hospital patients treated with supplements. injury or illness, unmasking glucose intolerance in the The best results in terms of mortality were obtained previously glucose-tolerant. Enteral and parenteral nu- in those aged >75 years, in those taking >400 kcal trition are valid options in the malnourished elderly, of supplement daily, those in a poor general condi- both in the hospital and at home. Elderly patients share tion, and those who were initially undernourished. In most indications and complications with adult patients, long-stay patients (>4 weeks), however, there was a though perhaps a greater focus should be placed on tendency for even the initially well-nourished patients function and quality of life than on mortality. to become malnourished. This may be prevented by Out of hospital, the focus for good glycaemic control using supplements. In the above meta-analysis, there is to minimize any long-term diabetic complications. also was a trend towards a reduced length of stay in In hospital, the rationale for good glycaemic control is hospital for the intervention group. to ensure a metabolic environment that promotes the For most patients, supplements should be prescribed best possible immune activity and wound healing. Hy- for a limited period of time, and continued only after a perglycaemia has a detrimental effect on the immune positive effect has been documented. The compliance system, adversely affecting chemotaxis, granulocyte and acceptance of supplementation by patients may adhesion, phagocytosis, intracellular killing and com- become problematic. Supplements should be given be- plement function [68]. The optimal blood glucose level tween meals and occasionally before bedtime, in order for ill patients receiving nutritional support is unclear, 224 CH 15 NUTRITIONAL PERSPECTIVES: DIABETES IN OLDER PEOPLE but various reports have provided many different blood feeds, but this is really only suitable when the enteral glucose targets and approaches to avoid both hyper- tube is pre-pyloric. and hypoglycaemia. The specific targets for glycaemic Insulin is required for all patients with type 1 diabetes control for each patient must take into consideration and for those with type 2 diabetes with significant hy- some of the following variables: age, prognosis, aeti- perglycaemia or critical illness. The type of insulin regi- ology of the hyperglycaemia, level of consciousness, men used will be tailored to the particular circumstances severity of any infection, degree of metabolic stress of the patient, ranging from a continuous infusion to the and immune status. use of subcutaneous, intermittent, quick-acting insulin Although the metabolic consequences of diabetes on a background of once- or twice-daily long-acting are known to involve fat metabolism and to result insulin. It should be noted that if insulin is added to in a significant dyslipidaemia, there are few defined a parenteral nutrition bag, some will be adsorbed onto management targets for serum lipids when giving nu- the plastic of the bag and cannulas. tritional support to diabetic patients. The degree of dyslipidaemia is frequently disproportional to the de- 15.6.1 Is artificial nutritional support gree of hyperglycaemia, and requires monitoring and necessary? control which is separate from the diabetes. Some of Artificial nutritional support is indicated for those pa- the clinical consequences of hyperlipidaemia include: tients who are malnourished, or who would become impairment of the immune response, endothelial dys- malnourished if not treated in this way. In order to function, an increased tendency to develop a coagu- benefit, patients must be fed for seven days or more lopathy and an exacerbation of insulin resistance. [70]. The goals of nutritional support for patients with Each hospital should have clearly stated monitor- diabetes are to maintain or improve nutritional sta- ing protocols for nutritional support [66, 67]. The fre- tus, to promote wound healing, to optimize glycaemic quency of monitoring depends on the clinical situa- control and optimal lipid control, and to avoid either tion, and must be individualized. Patients with hyper- hyperglycaemia or hypoglycaemia. glycaemia, and those who are severely malnourished, should be monitored closely to prevent metabolic 15.6.2 Route of artificial support complications, including the potential life-threatening [5, 66, 67] re-feeding syndrome associated with profound electrolyte disturbances and fluid overload. During re- When a decision has been made that nutritional feeding there is a switch from fat to carbohydrate support is required, the optimal route must be metabolism with an increase in insulin release. Dur- determined. Enteral nutrition should be used whenever ing carbohydrate repletion, insulin- stimulated glucose possible, as it enjoys many advantages over parenteral uptake is accompanied by an increased cellular uptake nutrition, including economic considerations, the of potassium, phosphorus and water, and magnesium avoidance of infections associated with parenteral requirements will increase as a consequence of stimu- nutrition, and a more physiological impact on the lation the sodium-potassium adenosine triphosphatase intestinal bacterial milieu. (ATPase) pump [5]. Enteral nutrition is suited to patients with diabetes Medical management must be employed in the el- due to the more physiological delivery of nutrients. derly patient with diabetes on nutritional support to The first line of nutritional support is an oral diet, and achieve the objectives of glycaemic control. Oral hy- only if the patient is at risk of aspiration or cannot meet poglycaemic agents (OHAs) are relatively contraindi- their nutritional requirements orally should other routes cated during critical illness. Metformin is not usually be considered. The decision requires a consensus from suitable for ill patients because there is a formal con- the clinical team and, if possible, the agreement of the traindication for its use in patients undergoing any form patient and/or caregivers. The options are: oral diet; of imaging that requires contrast media. The OHAs nutritional oral supplements; and tube feeding, whether most suited to hospital use are the short-acting in- pre-pyloric or post-pyloric. sulin secretagogues, such as repaglinide or nateglinide. When enteral nutrition is prolonged due to Long-acting sulphonylureas should be avoided as these persistent anorexia or dysphagia, then percutaneous are a potent cause of hypoglycaemia, particularly in the endoscopic gastrostomy (PEG) will often be the elderly. If necessary, OHAs may be uses in clinically route of choice for artificial nutrition in the elderly. stable patients receiving enteral (gastric administration) For this, three groups of patients can be identified: 15.7 ENTERAL TUBE FEEDING [43, 69] 225 (i) diabetes patients who will need prolonged home and gums, from a mixed or single source. The ratio enteral nutrition, probably due to persistent dysphagia of soluble to insoluble fibre in the mixed fibre source after a resolute disease; (ii) those who will obtain feeds varies, with some products having 50%/50% a short-term benefit before resuming oral nutrition, proportions while others contain 75% insoluble and such as with secondary anorexia after stress; and (iii) 25% soluble fibre. those who will die while on home enteral nutrition, • Specialist feeds: for the management of patients with due to their primary disease. In these patients enteral special needs, additional feeds are also available. nutrition can be considered as palliative. Typical patients will have renal failure, malabsorp- As might be expected, both life expectancy and the tion, electrolyte restrictions, milk protein intolerance health-related quality of life under nutritional support and inflammatory bowel disease. As a general rule, are poorer in elderly patients compared to younger elemental or semi-elemental feeds have an osmolar- patients. This is true for survival in home enteral − ity between 300 and 500 mOs l 1. nutrition patients [5], and after procedures such as PEG [71]. For obvious ethical reasons, no study has been designed to demonstrate any benefit of artificial 15.7.1 Composition of specialist feeds for nutrition versus an absence of nutritional support in the management of hyperglycaemia comparable groups. Rather, the only studies performed not only provided conflicting results but were also Most of the evidence used to support the use of spe- either observational or had non-comparable groups. cialized enteral feeds in diabetic management has been extrapolated from the general diabetic literature, and is aimed at avoiding hyperglycaemia. Specialist feeds for 15.7 Enteral tube feeding [43, 69] people with diabetes are available, and aim to reduce the usual high liquid carbohydrate content of stan- Enteral tube feeding (ETF) may be continuous, inter- dard feeds (>50% of calories as carbohydrate). A high mittent or overnight. When the regimen to be applied liquid carbohydrate content tends to exacerbate hyper- has been decided, the type of enteral tube feed must glycaemia, and often necessitates insulin therapy. Tube be selected, based on the nutritional and fluid require- feeding is associated with a more rapid increase in ments. The approximate composition of generic feeds postprandial glucose than solid diets of similar nutri- are as follows: tional composition. High postprandial glucose levels predispose to hypertriglyceridaemia [72]. − • Standard enteral tube feeds (1 kcal ml 1; osmolarity Compared with standard formulas, diabetes-specific − 201–250 mOs l 1): these contain 15–16% of energy formulas are typically higher in fat (40–50% of energy, as whole protein (milk protein-casein), 30–35% of with a large contribution from MUFAs, e.g., >60% energy as a mixture of long- and medium-chain fats of fat), with a lower carbohydrate content (35–40% [40% MUFA, 30% short-chain fatty acid (SFA) and of energy) and up to 15% of energy from fructose. poly-unsaturated fatty acids (PUFA)], such as lin- These nutrients could facilitate glycaemic management seed, sunflower, safflower or rapeseed oil, and may by delaying gastric emptying (fat and fibre), delaying also contain fish oil. Carbohydrates provide 50–56% the intestinal absorption of carbohydrate (fibre), and of the energy content of the feed (mainly present producing smaller glycaemic responses (fructose). A as maltodextrins), but may also contain sucrose, high proportion of MUFAs may also have beneficial oligosaccharides, polysaccharides, corn syrups and effects on lipid profiles. starches. Only short-term studies have been fulfilled using − specialized oral diets in which the carbohydrate content • High-energy feeds (1.5 kcal ml 1; osmolarity 300 − is reduced by increasing the MUFA content. These mOs l 1): these have the same percentage energy have been undertaken either as single test meals or over from macronutrients as the standard formula. The os- − short periods of time, and have involved relatively few molarity of these products is increased to 300 mOs l 1 subjects. due to the reduced volume of the product. Many of the nutrients included in tube feeds must • Fibre feeds: the amount of fibre per 100 ml is usually be chemically modified to enable delivery from a tube. between 1–2 g. The type of fibre ranges from soy, As the glycaemic response of a food is dependent on inulin wheat fibre, fructo-oligosaccharides, oat fibre its physical properties, changing nutrients from a solid 226 CH 15 NUTRITIONAL PERSPECTIVES: DIABETES IN OLDER PEOPLE phase to a liquid phase can radically alter the glycaemic Very few long-term studies have been reported exam- properties. With respect to glycaemic control, while ining clinical outcome. One study of ETF [73] showed there is good evidence of the beneficial effect of fibre the diabetes-specific formula to be associated with a in a solid diet, the addition of fibre to a liquid diet has trend towards a reduced incidence of pneumonia, fever not been shown to be of benefit [73]. In addition, fibre and urinary tract infection relative to the standard for- supplementation to tube feeding can be problematic, as mula. This may have clinical relevance for those hyper- optimal fibre blends increase the feed viscosity, which glycaemic patients who are at increased risk of infec- makes the formula flow through fine-bore feeding tubes tions. Further common comorbidities in patients with extremely difficult. Indeed, the biophysical properties diabetes include CVD and hyperlipidaemia. Although of a fibre in a liquid may be the reason why there diabetes-specific feeds had a higher fat content than has been a lack of improvement in glycaemic control standard feeds, it was suggested that diabetes-specific with tube feeds containing fibre. For tube feeds, the formulas had no detrimental effect on total cholesterol, postprandial insulin and glucose responses are related HDL-cholesterol or triglycerides. For ETF studies, it to the carbohydrate load, and not to its fibre content. was impossible to evaluate how far the administration The meta-analyses of Pohl and colleagues, based of the feeds might have influenced the metabolic ef- on studies of patients of medium age (70 years), with fects. A further consideration was the amount of feed insulin-treated type 2 diabetes, HbA1C <7.0% and an administered, as patients receiving ONSs may obtain indication for tube feeding due to dysphagia caused only 25% of their daily energy from this source, com- by neurological disorders, have shown that the use pared with up to 100% for tube-fed patients. of diabetes-specific oral and tube formulas (contain- National organizations [41, 42] generally rec- ing high proportions of MUFAs, fructose and fibre) ommend low-fat (25–35% of energy) and high- are associated with improved glycaemic control com- carbohydrate diets (45–60%), rich in complex pared to standard formulas. This shows that the use of carbohydrates, for those with diabetes. The situation diabetes-specific formulas, given either as oral nutri- for MUFAs is less clear, with the American Diabetes tional supports (ONSs) or enteral tube feeding (ETF), Association reporting that there is a lack of evidence results in a significantly lower postprandial rise in that MUFAs exert any long-term effects on glucose blood glucose, peak blood glucose concentrations and control or other metabolic parameters [41]. Formulas the glucose-versus-time area under the curve (AUC) that have a particularly high proportion of fructose in patients with diabetes. This was achieved with- should probably be administered with some caution to out any evidence of hypoglycaemia, and suggests that critically ill patients, who are at risk of lactic acidosis. glycaemic control may be facilitated by the use of Dietary therapy or enteral tube feeding, when given diabetes-specific enteral formulas compared to stan- under medical supervision, can be individualized to in- dard formulas in patients with diabetes. clude a more liberal use of fat (e.g. MUFAs), which The current meta-analyses found that the postpran- is particularly important in the treatment of malnour- dial rise in glucose concentration was lower, and ished patients to increase their dietary energy intake the peak glucose concentration reduced, following [75]. There is clearly a need for further research to diabetes-specific compared to standard formulas. Re- determine the role of enteral nutritional support and cent studies have also demonstrated a strong cor- diabetes-specific formulas on the management, clinical relation between postprandial glucose regulation and outcome and quality of life of malnourished patients cardiovascular complications in patients with diabetes, with diabetes. impaired glucose tolerance, and all-cause mortality, whereas no such correlation was demonstrated for 15.7.2 Complications of enteral nutritional fasting glucose control [74]. This implies that, by support improving glycaemic control, the long-term use of diabetes-specific versus standard enteral formulas may Gastrointestinal symptoms are the most frequent reduce cardiovascular complications in patients with byproducts of tube feeding, with gastroparesis and diabetes, although this proposal was not assessed by diarrhoea being the most common complications. the studies reviewed. Gastroparesis is extremely common among patients In some studies, diabetes-specific formulas reduced with diabetes and affects 30–75% of all patients un- the quantity of hypoglycaemic medication, and in some dergoing nutritional enteral feeding [5]. Gastroparesis cases also prevented the need for insulin injections. reduces the tolerance to enteral nutritional support, as 15.8 PARENTERAL NUTRITION [43, 69] 227 well as causing bloating, satiety, nausea and vomiting. non-protein energy from carbohydrate and fat). The irregular and unpredictable rate of gastric emp- The inclusion of fat improves substrate utilization, tying associated with gastroparesis can result in poor enables the delivery of fat-soluble vitamins, and glycaemic control, and also can cause an exacerbation also reduces the osmolarity of feeds which may be of gastroparesis. In addition to changing the enteral used for simultaneous peripheral feeding. The protein formula, a number of prokinetic drugs are available component in PN is made up of essential amino acids that can improve gastric emptying. If these strategies and soluble, non-essential amino acids. are unsuccessful, then a change to jejunal feeding may Parenteral nutrition is usually applied in an be helpful. all-in-one bag. A variety of PN pre-compounded Chronic diarrhoea occurs in 20–85% of people with bags are available, designed to meet the nutritional diabetes receiving enteral feeding, and can be a difficult requirements for most patients but, if possible, some management problem because it demands a system- hospital pharmacies are able to compound PN bags atic approach, including an awareness of the patient’s for individual patients. Parenteral nutrition should bowel history and any altered bowel habits prior to tube preferably be administered continuously over a 24 h feeding. It is important to consider all possible con- period, using a suitable infusion pump to minimize tributory factors for the diarrhoea, to take note of all infusion errors, and prevent marked swings in blood prescribed and non-prescribed medications being taken glucose and electrolyte values or rapid changes in (notably broad-spectrum antibiotics), and to consider fluid balance. any bacterial overgrowth and specific infections (e.g. Individuals with stress-induced glucose intolerance, Clostridium difficile) or other bowel pathology. It must and all diabetic patients, will require insulin during also be borne in mind that enteral feeding itself may be PN administration for glycaemic control, and all oral a cause of diarrhoea, due to the use of hyperosmolar hypoglycaemic agents must be stopped. The glycaemic feeds or feeds with an inadequate sodium content, as management during parenteral nutrition must be tai- well as rapid administration (e.g. bolus feeding). lored and adapted to the nutritional support required, taking into consideration that blood glucose levels reflect the underlying illness rather than the route of 15.8 Parenteral nutrition [43, 69] nutritional support [77]. Parenteral nutrition support can be optimized to Parenteral nutrition (PN) provides no added value over minimize both hyperglycaemia and hypoglycaemia by enteral feeding in patients with a functioning gastroin- adhering to the following points: testinal tract, and is related to an increased risk of complications and health costs [76]. Parenteral nutrition 1. Prevent overfeeding. is only indicated when enteral nutrition is contraindi- 2. Ensure that the infusion rate of carbohydrate in cated, which usually occurs when the gastrointestinal the PN does not exceed the patient’s glucose ox- tract is either non-functioning or not accessible. idation rate (6–7 mg kg−1 min−1), because exceed- Parenteral nutrition, for example, can result in ap- ing this value may increase the metabolic rate and proximately 30% of patients developing transient di- worsen glucose tolerance [78]. Alternative sugars abetes, and at least 15% developing hyperglycaemia have been tried experimentally as potential carbohy- [77]. Intravenous catheter-related infections are five- drate substitutes, including fructose, sorbitol, xylitol fold more prevalent in patients receiving central PN, a and glycerol, but have not successfully prevented or figure which is higher still in the presence of hyper- improved the hyperglycaemia. glycaemia. Parenteral nutrition is hyperosmolar and requires 3. Optimize the fat to carbohydrate ratio. Some groups access via a large central vein. Central access can have advocated increasing the fat component to be achieved either by a peripherally inserted central 60–70% of non-protein energy in PN in order to catheter (PICC) threaded up into a larger central vein reduce the carbohydrate component to 30–40%. A (which is more suitable for short-term PN), or by direct high fat content in PN increases the possibility of access to a central vein (PICC <15 cm; Hickman line hyperlipidaemia, and this has the potential in sep- or Portacath, as long-term lines). tic or critically ill patients to precipitate pancreatitis The energy content in PN is provided by a and renal failure. Reducing the carbohydrate com- mixed source of fat and carbohydrate (usually 50% ponent in patients with hyperglycaemia will reduce 228 CH 15 NUTRITIONAL PERSPECTIVES: DIABETES IN OLDER PEOPLE the glucose load of the feed, but this may not be incontinence. Any one of these can adversely affect the sufficient to allow the withdrawal of insulin. quality of life and result in a vicious cycle of decline, deterioration in self-management, greater demands on 4. Reduce the rate of the PN to prevent rebound caregivers, loss of independence and possible institu- hypoglycaemia before stopping PN. tionalization. An inadequate intake of energy and nutrients is a 15.9 A specific nutrition support common problem in demented patients. Such undernutrition may be caused by several factors, including formula for elderly patients? anorexia (commonly caused by polypharmacotherapy), insufficient oral intake (forgetting to eat), depression, There is no evidence in favour of a specific formula apraxia of eating or, less often, an enhanced energy in EN or PN in the elderly. Recently, Rees et al. [79] requirement due to hyperactivity (constant pacing). In have proved high-energy, high-protein EN diets capa- the advanced stages of dementia, dysphagia may de- ble of reaching a positive nitrogen balance very rapidly, velop, and this might be an indication for EN in a few which may be valuable in stressed elderly patients. cases. Enteral nutrition may be recommended during However, sodium reabsorption is lower and the thresh- the early stages of the disease, or after an acute weight old for thirst higher in elderly subjects, which high- loss in patients with Alzheimer’s disease. For patients lights the needs for water intake (30 ml kg−1 per day), with terminal dementia, EN is not recommended how- and this should be considered in the prescription of ever (grade of recommendation C) [66]. EN/PN formulas. Semi-elemental EN formulas are not preferable to their polymeric counterparts. Finally, fibre supplementation is able to improve bowel function 15.10.2 Neurological dysphagia with a reduced stool frequency and a more solid stool Nutritional therapy depends on the type and extent of consistency, without affecting the nutritional efficiency the swallowing disorder. This may range from normal of enteral feeding in hospitalized geriatric patients [80]. food, to mushy meals (modified consistency), thickened liquids of different consistencies or to total EN delivered via a nasogastric tube or PEG. In a Cochrane 15.10 Ethical issues analysis of interventions for dysphagia in acute stroke, EN delivered via PEG was associated with a greater Ethical issues are crucial in deciding when to start improvement of nutritional status compared to EN an elderly patient on artificial nutrition, there hav- delivered via a nasogastric tube [81]. As dysphagia ing been much recent public controversy regarding will rarely improve after two weeks, should severe life-sustaining technologies for elderly people. The pa- dysphagia persist for longer than 14 days after the tient’s informed consent is essential, although a family acute event, then a PEG should be placed immedi- member or a caregiver may act as possible surrogates. ately [66]. The decision must always be based on evidence, but if this is unavailable then the patient’s wishes (or those of his/her family) must be taken into account. In a 15.10.3 Pressure sores and diabetic foot therapeutic project, even if artificial nutrition may be Malnutrition increases the risk for the occurrence of withheld, it best not to do so. pressure sores, which are associated with an increased 15.10.1 Geriatric syndromes and nutrition risk of morbidity and mortality. Pressure sores develop in 4–10% of newly hospitalized patients, increasing in the older diabetic patient to 14% in long-term elderly care. Patients with dia- In contributing to their generally poorer quality of betes are a vulnerable group with poor wound healing. life, older persons with diabetes have higher rates To date, there is insufficient evidence to support the of CHD, hypertension and stroke, and are also at routine supplementation of micronutrients for wound greater risk for developing one of the common – but or leg ulcer healing using either multivitamins or vi- frequently overlooked – geriatric syndromes than their tamin C, with or without zinc. An improved healing age-matched, non-diabetic counterparts. These syn- of leg ulcers and wounds has been reported follow- dromes include depression, cognitive impairment, ining a 3-month period of zinc supplementation, given jurious falls, polypharmacy, persistent pain and urinary as 70 mg of zinc three times each day. A review on 15.11 CONCLUSIONS 229 nutrition and wound healing concluded that, while the taking into consideration their food preferences, their supplementation of hospitalized patients with zinc and cultural background, financial resources and support vitamin C may be reasonable, the routine use of vi- systems. A series of simple recommendations, drafted tamin C supplementation alone was unlikely to be with the help of the patient or their caregiver, would beneficial [56, 82]. more likely be successful. The elderly are often at risk Low protein and energy intake, BMI and albu- of nutritional deficiency and malnutrition, and the pres- minaemia are all risk factors for the development of ence of a chronic disease such as diabetes profoundly pressure sores in elderly patients. Additionally, oral affects their metabolism, placing them at still higher nutritional supplements could significantly reduce the risk. Unfortunately, our present understanding of this incidence of pressure ulcer development in at-risk pa- problem is somewhat limited, and additional research tients (Odds ratio 0.75, 95% CI: 0.62–0.89) [82, 83]. in this area is clearly required, mainly in order to ame- As with the effect of nutritional status on the healing of liorate the quality of life of the elderly diabetic. existing pressure ulcers, the scarce amount of available data suggests that malnutrition slows the healing process, and that an increase in protein and energy intake References raises the rate of healing. A systematic review by Strat- ton et al. showed that enteral nutritional support may 1. Narayan KM, Boyle JP, Thompson J et al. Lifetime significantly reduce (by 25%) the risk of developing risk for diabetes mellitus in the United States. JAMA pressure ulcers. [84]. 2003; 290: 1884–90. 2. Gu K, Cowie CC and Harris MI. Mortality in adults with and without diabetes in a national cohort of the 15.10.4 Oral health U.S. population, 1971–1993. Diabetes Care 1998; 21: Diabetes adversely affects oral health, increasing the 1138–45. risk of gingivitis and other oral infections. Gingivitis 3. Meneilly GS and Tessier D. Diabetes in elderly adults. is a major cause of tooth loss and pain that can affect J Gerontol A: Biol Sci Med Sci 2001; 56: M5–13. 4. Euronut – SENECA. Nutrition and the elderly in Eu- oral intake. Poor oral and dental health is linked with rope. 1st European Congress on Nutrition and Health chewing difficulties that can cause malnutrition, poor in the Elderly. The Netherlands, December. 1991 Eur general health and a reduced quality of life. There are J Clin Nutr 1991; 45 (Suppl 3). 1–196. dietary implications for those with no teeth or partial 5. Peake H. (2003) Inpatient nutritional support of sick dentures, as difficulties in eating can lead to a reduction patients with diabetes. In: G. Frost, A. Dornhorst and in the variety of food choices and an overall reduction R. Moses (eds). Nutritional Management of Diabetes in nutrient intake. Full dentures can cause a reduction Mellitus. John Wiley & Sons, Ltd. England. 2003; in food consumption due to the mouth feeling full, to a chapter 14; pp 215–229. greater time needed to eat, to causing embarrassment, 6. Morley JE. Anorexia of aging: physiologic and patho- and to changes in food flavours [85]. All patients logic. Am J Clin Nutr 1997; 66 (4): 760–73. should be encouraged to maintain good oral hygiene, 7. Newman AB, Arnold AM, Burke GL et al. Cardiovas- with special attention given to those with dry mouths cular disease and mortality in older adults with small or who eat more frequently due to a small appetite abdominal aortic aneurysms detected by ultrasonogra- or, in the case of a patient with type 1 diabetes, to a phy: The Cardiovascular Health Study. Ann Intern Med need for frequent snacks. Dental advice is required for 2001; 134 (3): 182–90. 8. Flodin L, Svensson S and Cederholm T. Body mass patients with chewing difficulties, pain and other oral index as a predictor of 1 year mortality in geriatric health problems. patients. Clin Nutr 2000; 19: 121–5. 9. Morley JE and Perry HM, III. The management of diabetes mellitus in older individuals. Drugs 1991; 41: 15.11 Conclusions 548–65. 10. Beaufrere B and Morio B. Fat and protein redistribution The dietary treatment of diabetes has long been the cor- with aging: metabolic considerations. Eur J Clin Nutr nerstone of management of this common disease, espe- 2000; 54 (Suppl. 3): S48–53. cially in the elderly diabetic, who may present a major 11. Morley JE, Thomas DR and Wilson MMG. Cachexia: challenge to the physician. Nutritional recommenda- pathophysiology and clinical relevance. Am J Clin Nutr tions for these patients should be highly individualized, 2006; 83: 735–43. 230 CH 15 NUTRITIONAL PERSPECTIVES: DIABETES IN OLDER PEOPLE

12. Janssen I, Baumgartner RN, Ross R et al. Skeletal 28. Fontaine KR and Allison DB. Does intentional weight muscle cutpoints associated with elevated physical dis- loss affect mortality rate? Eat Behav 2001; 2 (2): ability risk in older men and women. Am J Epidemiol 87–95. 2004; 159 (4): 413–21. 29. Kondrup J, Allison SP, Elia M, Vellas B and Plauth M. 13. Campbell AJ, Spears GF, Brown JS et al. Anthropo- ESPEN Guidelines for Nutrition Screening 2002. Clin metric measurements as predictors of mortality in a Nutr 2003; 22: 415–21. community population aged 70 years and over. Age 30. Fuhrman MP, Charney P and Mueller CM. Hepatic Ageing 1990; 19 (2): 131–5. proteins and nutrition assessment. J Am Diet Assoc 14. Morley JE and Silver AJ. Nutritional issues in nursing 2004; 104 (8): 1258–64. home care. Ann Intern Med 1995; 123 (11): 850–9. 31. Guigoz Y, Lauque S and Vellas BJ. Identifying the 15. Cederholm T, Jagren C and Hellstrom K. Outcome of elderly at risk for malnutrition. The Mini Nutritional protein-energy malnutrition in elderly medical patients. Assessment. Clin Geriatr Med 2002; 18 (4): 737–57. Am J Med 1995; 98 (1): 67–74. (Available at: http://mna-elderly.com) 16. Herrmann FR, Safran C, Levkoff SE and Minaker KL. 32. Bauer JM, Vogl T, Wicklein S, Trogner¨ J, Mohlberg¨ W Serum albumin level on admission as a predictor of and Sieber CC. Comparison of Mini Nutritional Assess- death, length of stay, and readmission. Arch Intern Med ment, Subjective Global Assessment and Nutritional 1992; 152: 125–30. 17. Sullivan DH, Sun S and Walls RC. Protein-energy Risk Screening (NRS 2002) for Nutritional Screen- undernutrition among elderly hospitalized patients: a ing and Assessment in Geriatric Hospital Patients. Z prospective study. JAMA 1999; 281: 2013–19. Gerontol Geriat 2005; 38: 322–7. 18. Calle EE, Thun MJ, Petrelli JM et al. Body-mass index 33. Kondrup J, Rasmussen HH, Hamberg O, Stanga Z, and and mortality in a prospective cohort of US adults. N the Ad Hoc ESPEN Working Group. (2003) Nutritional Engl J Med 1999; 341 (15): 1097–105. risk screening (NRS 2002): a new method based on an 19. Potter JF, Schafer DF and Bohi RL. In-hospital mortal- analysis of controlled clinical trials. Clin Nutr 2003; ity as a function of body mass index: an age-dependent 22: 321–36. variable. J Gerontol 1988; 43: M59–63. 34. Kondrup J, Allison SP, Elia M, Vellas B and Plauth 20. Roberts SB, Fuss P, Heyman MB et al. Control of food M. (2003) ESPEN Guidelines for Nutrition Screening intake in older men. JAMA 1994; 272 (20): 1601–6. 2002. Clin Nutr 2003; 22: 415–21. 21. Walker D and Beauchene RE. The relationship of lone- 35. Morley JE. Pathophysiology of anorexia. Clin Geriatr liness, social isolation, and physical health to dietary Med 2002; 18: 661–3. adequacy of independently living elderly. J Am Diet 36. Reilly HM. Screening for nutritional risk. Proc Nutr Assoc 1991; 91 (3): 300–4. Soc 1996; 55: 841–53. 22. Thomas P, Hazif-Thomas C and Clement JP. Inﬂu- 37. Malnutrition Advisory Group (MAG). MAG ence of antidepressant therapies on weight and ap- guidelines for detection and management of mal- petite in the elderly. J Nutr Health Aging 2003; 7 (3): nutrition. 2000, Redditch, UK: British association 166–70. for Parenteral and Enteral Nutrition. (Available 23. Wilson MM, Vaswani S, Liu D et al. Prevalence and at:www.bapen.org.uk/must tool.html) causes of undernutrition in medical outpatients. Am J 38. Detsky AS, McLaughlin JR, Baker JP, Johnston S, Med 1998; 104 (1): 56–63. Whittake S, Mendlson RA and Jeejeebhoy KN. What 24. Flegal KM, Carroll MD, Ogden CL et al. Prevalence is subjective global assessment of nutritional status? and trends in obesity among US adults, 1999–2000. JPEN. 1987; 11: 8–13. JAMA 2002; 288 (14): 1723–7. 39. Buzby GP, et al. Prognostic nutritional index in gas- 25. Heiat A, Vaccarino V and Krumholz HM. An evidence-based assessment of federal guidelines for trointestinal surgery. Am J Surg 1980; 139: 160–167. overweight and obesity as they apply to elderly 40. Omran ML and Salem P. Diagnosing undernutrition. persons. Arch Intern Med 2001; 161 (9): 1194–203. Clin Geriatr Med 2002; 18: 719–36. 26. Villareal DT, Apovian CM, Kushner RF et al. Obesity 41. American Diabetes Association. (2007) Nutrition Rec- in older adults: technical review and position statement ommendations and Interventions for Diabetes. A posi- of the American Society for Nutrition and NAASO, tion statement of the American Diabetes Association. The Obesity Society. Am J Clin Nutr 2005; 82 (5): Diabetes Care 2007; 30 (Suppl. 1): S48–65. 923–34. 42. IDF (Europe). A desktop guide to Type 2 diabetes 27. Meneilly GS. Diabetes in the elderly. Med Clin N Am mellitus. European Diabetes Policy Group, Diabet Med 2006 (90): 909–23. 1999; 16 (9): 716–30. 15.11 CONCLUSIONS 231

43. NICE guidelines (2006). Nutrition support in adults: falls? Results of a randomized trial. J Am Geriatr Soc Oral nutrition support, enteral tube feeding and par- 2005; 53 (11): 1881–8. enteral nutrition. Published by the National Collabo- 58. Hanley DA and Davison KS. Vitamin D insufficiency rating Centre for Acute Care at The Royal College of in North America. J Nutr 2005; 135 (2): 332–7. Surgeons of England. (Available at: www.rcseng.ac.uk 59. Hollis BW. Circulating 25-hydroxyvitamin D levels or www.nice.org.uk). indicative of vitamin D sufficiency: implications for 44. Milton JE, Briche B, Brown IJ, Hickson M, Robertson establishing a new effective dietary intake recommen- CE and Frost GS. Relationship of glycaemic index with dation for vitamin D. J Nutr. 2005; 135 (2): 317–22. cardiovascular risk factors: analysis of the National 60. Ervin RB and Kennedy-Stephenson J. Mineral intakes Diet and Nutrition Survey for people aged 65 and older. of elderly adult supplement and non supplement users Public Health Nutr. 2007; 10 (11): 1321–35. in the third National Health and Nutrition Examination 45. Finch S, Doyle W, Lowe C, Bates CJ, Prentice A, Survey. J Nutr 2002; 132 (11): 3422–7. Smithers G and Clarke P. National Diet and Nutrition 61. Grant AM, Avenell A, Campbell MK et al.Oralvi- Survey: people aged 65 years and over. Vol. 1. Report tamin D3 and calcium for secondary prevention of of the Diet and Nutrition Survey. The Stationary Office, low-trauma fractures in elderly people (Randomised London, 1998. Evaluation of Calcium Or vitamin D, RECORD): a 46. Chapman IM. Nutritional disorders in the elderly. Med randomised placebo-controlled trial. Lancet 2005; 365 Clin N Am 2006; 90 (5): 887–907. (9471): 1621–8. 47. Hickson M. Malnutrition and ageing. Postgrad Med J 62. Andres E, Loukili NH, Noel E et al. Vitamin B12 2006; 82: 2–8. (cobalamin) deficiency in elderly patients. Can Med 48. Nitzke S, Freeland-Graves J and the American Dietetic Assoc J 2004; 171 (3): 251–9. Association. (2007) Position of the American Dietetic 63. Wald DS, Law M and Morris JK. Homocysteine and Association: total diet approach to communicating food cardiovascular disease: evidence on causality from a and nutrition information. J Am Diet Assoc. 2007; 107 meta-analysis. Br Med J 2002; 325 (7374): 1202. (7): 1224–32. 64. Rampersaud GC, Kauwell GP and Bailey LB. Folate: 49. Zizza CA, Tayie FA and Lino M. Benefits of snacking a key to optimizing health and reducing disease risk in in older Americans. J Am Diet Assoc. 2007; 107 (5): the elderly. J Am Coll Nutr 2003; 22 (1): 1–8. 800–6. 65. Joshi S and Morley JE. Vitamins and minerals in the 50. Hickson M and Wright L. (2003) Nutritional manage- elderly. In: MSJ Pathy, AJ Sinclair and JE Morley ment of the elderly person with diabetes. In: G. Frost, (eds). Principles and practice of geriatric medicine.4th A. Dornhorst and R. Moses (eds). Nutritional Manage- ed. John Wiley & Sons Ltd. Chichester, England. 2006; ment of Diabetes Mellitus. John Wiley & Sons, Ltd. Vol. 1, pp. 329–46. England. 2003; chapter 10; pp 147–168. 66. Volkert D, Berner YN, Berry E, Cederholm T, Coti 51. American College of Sports Medicine Position Stand. Bertrand P, Milne A, Palmblad J, Schneider S, Sobotka (1998) Exercise and physical activity for older adults. S, Stanga Z, DGEM (German Society for Nutritional Med Sci Sports Ex 1998; 30: 992–1008. Medicine), Lenzen-Grossimlinghaus R, Krys U, Pirlich 52. Gregg EW, Gerzoff RB, Caspersen CJ et al. Relation- M, Herbst B, SchutzT,Schr¨ oer¨ W, Weinrebe W, ship of walking to mortality among US adults with Ockenga J, Lochs H. ESPEN Guidelines on Enteral diabetes. Arch Intern Med 2003; 163: 1440–7. Nutrition: Geriatrics. Clin Nutr 2006; 25(2): 330–60. 53. Physical Activity Guidelines for Americans (2008). 67. Milne AC, Avenell A and Potter J. Meta-analysis: Chapter 5: Active older adults. Available at: www. Protein and energy supplementation in older people. health.gov/paguidelines/guidelines/chapter5.aspx. Ava- Ann Intern Med 2006; 144: 37–48. ilable as pdf-document at: www.health.gov/paguide- 68. Pomposelli JJ, Baxter JK, Babineau TJ, Pomfret EA, lines/pdf/paguide.pdf. Driscoll DF, Forse RA and Bistrain BR. Early post 54. Waugh NR and Robertson AM. Protein restriction for operative glucose control predicts nosocomial infection diabetic renal disease. Cochrane Database Syst Rev rate in diabetes patients. J Parenter Enteral Nutr 1998; 2000; 2: CD002181. 22: 77–81. 55. Swade TF and Emanuele NV. Alcohol and diabetes. 69. Task Force of A.S.P.E.N; American Dietetic Associ- Compr Ther 1997; 23: 135–40. ation Dietitians in Nutrition Support Dietetic Practice 56. Mooradian A, Failla M, Hoogwerf BJ, Maryniuk M Group, Russell M, Stieber M, Brantley S, Freeman AM, and Wylie-Rosett J. Selected vitamins and minerals in Lefton J, Malone AM, Roberts S, Skates J, Young LS; diabetes. Diabetes Care 1994; 17: 464–79. A.S.P.E.N. Board of Directors; ADA Quality Manage- 57. Flicker L, MacInnis RJ, Stein MS et al. Should older ment Committee. American Society for Parenteral and people in residential care receive vitamin D to prevent Enteral Nutrition (A.S.P.E.N.) and American Dietetic 232 CH 15 NUTRITIONAL PERSPECTIVES: DIABETES IN OLDER PEOPLE

Association (ADA): standards of practice and standards nutrition: a pragmatic study. Nutrition 2001; 17: of professional performance for registered dietitians 1–12. (generalist, specialty, and advanced) in nutrition sup- 77. Pitts DM, Kilo KA and Pontious SL. Nutritional support. Nutr Clin Pract. 2007; 22 (5): 558–86. port for the patient with diabetes. Crit Care Nurs Clin 70. Sandstrom R, Drott C, Hyltander A, Arfvidsson B, North Am 1993; 5: 47–56. Schersten T, Wickstrom I and Lundholm K. The effect 78. Orr ME. Hyperglycemia during nutritional support. Crit of postoperative intravenous feeding (TPN) on outcome Care Nurs 1992; 12: 64–70. following major surgery evaluated in a randomized 79. Rees RG, Cooper TM, Beetham R, Frost PG and study. Ann Surg 1993; 217: 185–95. Silk DB. Influence of energy and nitrogen contents 71. Schneider SM, Raina C, Pugliese P, Pouget I, Rampal of enteral diets on nitrogen balance: a double blind P and Hebuterne X. Outcome of patients treated with prospective controlled clinical trial. Gut. 1989; 30(1): home enteral nutrition. J Parenter Enteral Nutr. 2001; 123–9. 25: 203–9. 80. Vandewoude MF, Paridaens KM, Suy RA, Boone MA 72. Schrezenmeir J. Rationale for specialized nutrition sup- and Strobbe H. Fibre-supplemented tube feeding in the port for hyperglycaemic patients. Clin Nutr 1998; 17 hospitalised elderly. Age Ageing 2005; 34: 120–4. (Suppl. 2): 26–34. 81. Bath PM, Bath FJ and Smithard DG. Interventions for 73. Craig LD, Nicholson S, Silverstone FA and Kennedy dysphagia in acute stroke. Database Syst Rev. 2000: RD. Use of a reduced carbohydrate, modified-fat en- CD000323. teral formula for improving metabolic control and clin- 82. Collins C. A practical guide to nutrition and pressure ical outcomes in long-term care residents with type 2 sores. Compl Nutr 2001; 1: 25–7. diabetes: results of a pilot trial. Nutrition 1998; 14: 83. Guerin´ O, Andrieu S, Schneider SM et al. Different 529–34. modes of weight loss in Alzheimer disease: a prospec- 74. Heine RJ, Balkau B, Ceriello A, Del Prato S, Horton tive study of 395 patients. Am J Clin Nutr. 2005; 82: ES and Taskinen MR. What does postprandial hyper- 435–41. glycaemia mean? Diabet Med 2004, 21: 208–13. 84. Stratton RJ, Ek AC, Engfer M, Moore Z, Rigby P, 75. Elia M, Ceriello A, Laube H, Sinclair AJ, Engfer M Wolfe R and Elia M. Enteral nutritional support in pre- and Stratton R. Enteral nutritional support and use of vention and treatment of pressure ulcers: a systematic diabetes-specific formulas for patients with diabetes: review and meta-analysis. Ageing Res Rev 2005; 4(3): A systematic review and meta-analysis. Diabetes Care 422–50. 2005; 28: 2267–79. 85. Mojon P, Budtz-Jorgensen E and Rapin CH. Rela- 76. Woodcock NP, Zeigler D, Palmer D, Buckley P, tionship between oral health and nutrition in very old Mitchell CJ and MacFie J. Enteral versus parenteral people. Age Ageing 1999; 28: 463–8. 16 Early Management of Type 2 Diabetes

Alan Sinclair Bedfordshire and Hertfordshire Postgraduate Medical School, University of Bedfordshire, Putteridge Bury Campus, Luton, UK

people with diabetes is complex and often expensive, Key messages with 60% of total health care expenditure being related to hospitalizations [3]. In addition, informal car- • Many older people with diabetes have a set of ers (care-givers) may lose up to $20 000 per year in unique characteristics which must be considered lost employment [4]. when planning initial management. Although, type 2 diabetes is clearly not limited to • High circulating glucose levels in the elderly older people, the disease’s unique characteristics be- can produce a varied symptom and sign profile gin to surface in those subjects with this condition which is distinct from the common osmotic who are of advanced age (>75 years), may have ad- profile. • Achieving high-quality diabetes care can be real- ditional major comorbidities and multiple medications, ized by use of a five-step management protocol. and who have personal and diabetes care issues that may not be coordinated in the community [5, 6]. These characteristics are listed in Table 16.1. It is also impor- 16.1 Introduction tant to identify those patients who are ‘frail’ because their management will be modified, their aims of care Diabetes mellitus in ageing subjects is a chronic metabolic disorder associated with macrovascular disease, a spectrum of functional impairments and, in Table 16.1 Defining characteristics of older subjects many cases, premature death [1]. Some evidence of im- with diabetes. proved care has emerged during the decade-and-a-half • High levels of medical comorbidities since the plight of elderly people with diabetes was first • highlighted, when there was a call for interventional Age-related impairment of functional ability strategies to reduce early functional decline. The call • Increased vulnerability to hypoglycaemia was also made for an individual-based plan of care, • Overlapping and often limited medical follow-up by giving details of realistic targets of glycaemia, reg- primary-care physicians, and hospital-based specialists ular screening for complications, and involving both • A management system which involves spouses and inpatients and carers in educational programmes [2]. formal carers to a greater extent As with several other chronic disease states, care for

Diabetes in Old Age Third Edition Edited by Alan J. Sinclair © 2009 John Wiley & Sons Ltd. ISBN: 978-0-470-06562-4 234 CH 16 EARLY MANAGEMENT OF TYPE 2 DIABETES adjusted, and their prognosis affected. Frailty is likely illnesses, although with detailed enquiry directly rein the presence of significant physical or cognitive de- lated symptoms of diabetes can be confirmed in a large cline, care home residency, malnourishment, and when number of cases. Some patients do not have the clas- long-term vascular disease such as coronary artery sic features of either diabetic ketoacidosis (DKA) or disease (CAD) is present, which limits severely both hyperosmolar non-ketotic (HONK) coma, but present quality of life and well-being. The adjustment of both with a ‘mixed’ disturbance of hyperglycaemia (blood glycaemic and blood pressure targets for the presence glucose levels 15–25 mmol l−1, an arterial blood pH of frailty has been recommended [7, 8] although the ev- of 7.2–7.3 (i.e. not particularly acidotic), and without idence base to justify this is weak due to a lack of data. marked dehydration or any change in their level of consciousness. 16.2 Developing the case for The insidious presentation may delay diagnosis and high-quality diabetes care partially account for the high prevalence of diabetic complications at the time of diagnosis. A better screen- The presentation of diabetes in the older patient is var- ing for complications at the time of diagnosis is, there- ied, and the symptoms and signs profiles of those with fore, part of the rationale for promoting quality dia- hyperglycaemia are often unexpected (Table 16.2). betic care for older patients with diabetes (Table 16.3). Many cases are detected by noting hyperglycaemia dur- Other recommendations relating to metabolic control, ing hospital admissions for other morbidities or acute eye screening and specialist follow-up complement this

Table 16.2 The varying relationship between raised plasma glucose levels (>9 mmol l−1) and patient symptom and sign proﬁle. Lethargy Usually glucose >11 mmol l−1 Increased micturition Disturbed sleep patterns Increased fall rate Dehydration Incontinence Visual impairment Increased fall rate Poor mobility Erectile impotence Complicated by vascular disease Pain Limb pain and decreased threshold Cognitive impairment Memory disorder Psychomotor slowing Depressive symptoms Irritability and intolerance

Table 16.3 Rationale for high-quality diabetes care for older people.a Screening and early diagnosis may prevent progression of undetected vascular Level of evidence 2++, Grade of complications recommendation (C) Improved metabolic control will reduce cardiovascular risk Level of evidence 2++, Grade of recommendation (B) Improved screening for maculopathy and cataracts will reduce visual impairment Level of evidence 2+, Grade of and blind registrations recommendation (C) An integrated approach to management of peripheral vascular disease and foot Level of evidence 2++, Grade of disorders will reduce amputation rate recommendation (B) Improved primary care and specialist follow-up will reduce hospital admission rate Level of evidence 3, Grade of recommendation (D)

aAdapted from Ref. [8]. 16.3 AIMS IN THE EARLY MANAGEMENT 235 approach. Although few long-term studies have specif- to reducing vascular complications, maintaining func- ically involved older patients (e.g. aged >75 years) tional status and ensuring the highest quality of life and none has attempted to assess the benefits of inter- commensurate with effective care. vention in frail subjects, a number of potential benefits may accrue from enhancing metabolic control. These include the removal of fatigue, a reduced risk 16.3.1 Acute presentation of metabolic decompensation and admission to hospi- Diabetes in elderly patients with type 2 diabetes may tal, a reduced carer support, and the avoidance of early present acutely in several ways: in DKA or as HONK functional decline. coma or, more commonly, as hyperglycaemia without significant ketosis or increased osmolality with or without coexisting acute illness; for example, an acute cere- 16.3 Aims in the early management brovascular accident (a mixed metabolic disturbance). Various precipitating factors for HONK coma have The overall assessment schedule for an older patient been identified (Table 16.5), and patients may present with newly diagnosed diabetes should mimic that of a comatose due to the combined effects of marked hy- younger individual, and form part of an integrated plan perglycaemia (glucose levels >30 mmol l−1) and dehy- of diabetes care. This will involve a comprehensive dration (serum osmolality >650 mOsm). In all cases history, physical examination and laboratory work-up. of presentation in a coma, other non-diabetes-related The essential components of this are well described causes should be excluded, such as head injury, stroke, in the Standards of Medical Care document of the alcohol or drug overdose. American Diabetes Association [9]. Seriously ill patients require insulin therapy (espe- The essential aims of management are summarized cially those with ketones), given either as an intra- in Table 16.4, and reflect the approach that aims are not venous infusion or by regular subcutaneous injections fixed, vary with time, and have different perspectives. of short-acting insulin, complemented by intravenous In situations where life expectancy is estimated to be rehydration. In all severe cases, arterial blood gases at least 7 years, the aims of care should be tailored should be measured to assess the acid–base balance.

Table 16.4 Aims in managing diabetes in the elderly. 16.3.2 Non-acute presentation Health professional-oriented The majority of elderly type 2 diabetic patients are not • To reduce fatigue due to raised glucose levels severely unwell at presentation, and should ideally be • To assess the impact on management and patient out- managed in the community by an interested general come of coexisting disease (e.g. ischaemic heart disease, practitioner or member of a primary care team with peripheral vascular disease) some additional competencies in diabetes care. At this • To prevent undesirable weight loss and maintain nutri- stage there are four important objectives: (i) to satisfy tional well-being yourself that the patient has diabetes; (ii) to screen for • Avoid hypoglycaemia and other adverse drug reactions complications; (iii) to identify who will be responsible • To screen for and prevent complications • To reduce the risk of functional impairment and disability Table 16.5 Hyperosmolar non-ketotic coma (HONK): precipitating factors. • To achieve a normal life expectancy for patients where possible • 50% Unknown • Infection Patient (and family/carer)-oriented • Operation (surgery) • To maintain general social and health well-being and • good quality of life Myocardial infarction • • To acquire skills and knowledge and understanding to Stroke adapt to changing requirements in their lifestyle • Drugs: propranolol, thiazides • Avoid dependency and institutionalization • Steroids, dialysis, (glucose drinks) 236 CH 16 EARLY MANAGEMENT OF TYPE 2 DIABETES for diabetic care (i.e. the patient or somebody else); succession of less powerful lenses, a direct inspection and (iv) to initiate treatment. of the cornea, anterior chamber and lens is possible. Diabetic retinopathy should be looked for after pupil- Making the diagnosis lary dilation using 0.5–1.0% tropicamide eye-drops. Relative contraindications for this include those with A detailed discussion of the criteria used to make a previous eye surgery, lens implants or a history of diagnosis of diabetes can be found in Chapter 3. narrow-angle glaucoma. The precipitation of previ- It should be appreciated, however, that many pa- ously undiagnosed acute glaucoma, although distress- tients may not be able to provide an accurate history ing at the time, may be a service to the patient in of symptoms. Moreover, the ‘classic’ symptoms of the long run as treatment may prevent further visual polyuria and polydipsia due to excessive glycosuria loss. Patients with scattered microaneurysms and blot may be absent owing to the raised renal threshold found in elderly subjects. A true fasted blood sample is haemorrhages require review at 6 months. This part often difficult to obtain, and may be normal in any case of the examination may be aided by a red-free filter [10], while an oral glucose tolerance test (OGTT) may to improve blood vessel examination. Diabetic mac- be thought of as time-consuming and inconvenient. In ulopathy can be sight-threatening and requires urgent addition, many patients have elevated plasma glucose referral to the ophthalmologist. Other reasons for refer- levels which are secondary to acute illness, diabeto- ral include the presence of yellow, waxy hard exudates, genic drug therapy, or other stress-inducing disorders. proliferative retinopathy, severe cataract formation, or If the physician has any doubt about the diagnosis of rapid decrease in VA, for example within the previous diabetes, it is wise not to treat but to retest later and 3 months (noticed by the patient – subjective; or evi- to use an OGTT if necessary. dence of a two-line deterioration in VA using a Snellen chart – objective). Screening for complications at diagnosis Examination of the limbs for sensory neuropathy should include an assessment of knee and ankle re- A detailed history may reveal symptoms of a distal flexes, sensation by testing with a nylon monofilament sensory diabetic neuropathy, such as numbness, paraes- (e.g. 5.07 Semmes–Weinstein), pin-prick and cotton thesiae, burning pains and hyperaesthesiae from bed- wool, vibration sense by 128 Hz tuning fork (bearing clothes at night-time. Symptoms of postural hypoten- in mind the age-associated loss of vibration sense) and sion (especially after treatment for coexisting hyper- proprioception. Infection, foot ulceration, the presence tension with vasodilators has been started), diarrhoea of pressure areas and the presence of sharp, poorly cut or constipation and impotence should provide an alert nails requires referral to the podiatrist. Management to the possibility autonomic neuropathy. Symptoms may also include radiology, antibiotic therapy, rest, of claudication should be inquired about. A physical use of pressure-relieving devices, and even surgery. examination requires the measurement of lying and Effective education, including advice about suitable standing blood pressures, and an assessment of the pe- footwear, reduces the risk of new foot lesions [11] but ripheral blood vessels. Visual acuity (VA) can be checked using a 3 m a recent targeted approach to secondary prevention did Snellen chart; patients with a VA worse than 6/6 (US not demonstrate benefit [12]. 20/20) in either eye should be examined using the Other investigations include: serum creatinine, pinhole test, which will partially correct for a refractive glycosylated haemoglobin (HbA1c), lipid profile error. Alternatively, they may use their distance glasses [triglycerides, total and high-density lipoprotein (if worn). In patients with poor VA which remains (HDL) cholesterol] in those aged less than 75 years, unaltered or worsens in the pinhole test, the retina especially those with CAD. In patients who may should be closely inspected for lesions, particularly have had undiagnosed diabetes for some time with those of maculopathy. marked hyperglycaemia, hyperlipidaemia may be Direct ophthalmoscopy should start with the lens present. In these cases, it is worth rechecking the at zero and a red reflex obtained. When present, this patient’s lipid profile after 6 months to determine indicates that there is no significant evidence of a whether the treatment has reversed the abnormality. cataract, vitreous haemorrhage or retinal detachment. An electrocardiogram seeking ischaemia, arrhythmias By setting the lens at +10 D initially, and using a and ventricular hypertrophy is useful. 16.4 INITIAL TREATMENT 237

Urinalysis (in the absence of infection) may demon- availability of services, cultural and dietary differences, strate proteinuria, although this is not particularly com- it is important to tailor the educational package to meet mon at diagnosis in the elderly. Microalbuminuria is their needs and provide information about any regional also common at diagnosis, but this may occur sec- or national societies/organizations involved in diabetes ondary to hypertension or congestive heart failure. It care. is routinely screened for in the author’s diabetic clinic when the dipstick test is negative for protein. A Barthel scale and mini-mental state examination to assess both 16.4 Initial treatment physical disability and mental function should be completed [13]. Brain failure may make the patient totally In patients whose glucose values lie between 8 and − dependent on others (spouse, other relative or commu- 17 mmol l 1, and who are not troubled by symptoms, nity nurse) for both treatment and monitoring. an initial 6- to 12-week course of dietary instruction only is warranted. The main elements of a suitable Identifying responsibilities in diabetes dietary plan include consuming 50–55% of total energy management intake as carbohydrate (including a daily fibre intake of at least 30 g), 30–35% fat intake (<10% saturated In most situations, an individual care plan must be fat) and 10–15% protein. adopted and agreed by all concerned. This may be Dietetic treatment will depend on several factors, but organized by the primary care physician (general prac- must include the patient’s ability to cooperate, physical titioner), although diabetes specialist nurses can play and mental well-being, and their natural desire to be an important role in this decision-making. This will consist of identifying the principal informal carer, set- independent. This process is a form of negotiation, ting realistic glycaemic goals, planning the timing and and some dieticians are today developing a ‘Getting frequency of visits, and being aware of the indications Started’ diet sheet for initial management. There is a for hospital referral to a specialist (Table 16.6) or ad- shift away from traditional ‘food exchanges’ towards a mission. Ideally, the health care team should aim to more generalized plan of healthy eating (provision of provide written information about diabetes for each ‘healthy eating messages’). In those patients who are newly diagnosed patient (and informal carer where overweight, a plan of fat restriction may be beneficial. appropriate) and organize several educational tutorials Other practical advice, including alcohol consumption over the next 6–12 months. and the benefits of exercise, are often given at this Wherever possible, a multidisciplinary approach and time. philosophy shared with the patient (or informal carer) When dietary advice fails to reduce levels of gly- is recommended, with the promotion of self-advocacy caemia or improve patient well-being, or when initial − being an important goal. With diabetic elders from random glucose levels are greater than 14 mmol l 1 ethnic minority backgrounds, who may pose special and/or a patient feels unwell, then several treatment problems of language and communication, access and options are available (Table 16.7), although in most cases oral agents are then prescribed. Table 16.6 Indications for referral to hospital specialist for elderly patients with diabetes. Table 16.7 Treatment options for diet failures. • Patients with severe complications (e.g. maculopathy, • foot ulceration, peripheral vascular disease) Further period of intense dietary therapy requiring inputs from both physician and dietician • Patients whose metabolic symptom control is suboptimal • irrespective of treatment (e.g. oral agents or insulin) Specified and appropriate exercise programme • • Complex management problems in those with coexisting Guar gum: little used in UK clinical practice disease (e.g. patients with chronic pulmonary disease • Acarbose: an alpha-glucosidase inhibitor taking steroids) • Oral hypoglycaemic agents: sulphonylureas, metformin, • Patients with increasing dependency and immobility (e.g. a meglitinide, thiazolidinediones post-stroke) • Insulin therapy: usually considered on a temporary basis • Patients not adequately cared for in primary care in well- defined circumstances only, such as acute illness 238 CH 16 EARLY MANAGEMENT OF TYPE 2 DIABETES 16.4.1 Physical activity and exercise in turn increases the risk of cardiovascular disease. The term ‘syndrome X’ or metabolic syndrome en- Structured physical activity (exercise) as part of a compasses these relationships, which also includes dys- lifestyle intervention programme (as an adjunct to lipidaemia [17]. Third, treatment with sulphonylureas, proper diet and weight control) protects against the a meglitinide or insulin is associated with hyperin- development of type 2 diabetes [14]. Brisk walking sulinaemia, which may promote both weight gain and for 30 min each day, or swimming for 45–60 min up paradoxically increase insulin resistance. These factors to three times per week are appropriate for older peo- are important and should be considered when antidia- ple. Aerobic exercise of this nature has been associ- betic therapy is instituted. ated with varying improvements in glycaemia, weight, lipid profile and blood pressure [15]. Combined train- 16.4.2 Oral agents ing which also involves moderate intensity resistance training (two sets of 12–20 repetitions) on a regular In choosing a specific antidiabetic drug, several factors basis can also lead to increases in muscle strength and must be considered, including renal and hepatic func- increased glucose disposal rates [16]. Prolonged or un- tion, coexisting disease, possible drug interactions, and usual exercise may be harmful, however, especially the likelihood of producing significant hypoglycaemia. in those with underlying cardiovascular disease, those There are no oral agents in routine practice which receiving insulin therapy (because of delayed hypo- are not recommended in older people, but knowledge glycaemia), and those with sensory loss in their feet, of their varying clinical pharmacology and the effect where tissue damage may be sustained. of ageing can be helpful in optimizing patient safety. In general, exercise has several beneficial actions: For this reason, glibenclamide (glyburide) and chlor- the lowering of hyperinsulinaemia and an improvement propamide, both of which have prolonged durations of in glucose tolerance occurs probably secondary to action, can accumulate in renal dysfunction, and have a a reduction in insulin resistance. The lipid profile high associated risk of hypoglycaemia, sometimes with becomes less atherogenic, with a reduction in total fatal consequences [18]. Hence, they should not be pre- plasma cholesterol and triglycerides while increasing scribed for diabetic subjects aged ≥60 years. Patients HDL-cholesterol. A fall in blood pressure may occur should be warned of the possibility of hypoglycaemia not only as a direct result of exercise but also as developing, and educated with practical advice on how the effect of weight loss. Some patients (especially to both avoid and prevent this potentially serious situa- older patients) are unable to participate in exercise tion developing. In relatively newly diagnosed patients, programmes because of decreased joint mobility due a failure to achieve acceptable glycaemic targets with to diabetes-related joint stiffness and/or osteoarthritis, diet and a single antidiabetic agent (e.g. a sulphony- or because of a previous stroke. The age-related loss lurea) after 6 months should lead to a further review of muscle mass seen in adults aged ≥50 years can of treatment. be pronounced in those in their 70s or 80s. Moreover, The International Diabetes Federation (European re- when this is combined with reduced energy expenditure gion) has published guidelines of diabetes care for type and decreased exercise levels, it often leads to an 2 diabetes [19], and a new updated version will be increase in weight, skeletal muscle fat deposition, and available in 2009. In the current guidance, no specific reduced lower limb muscle power. Limited exercise stepwise algorithm has been adopted for drug treat- only is possible in those with poor metabolic control ment, leaving the choice to the individual practitioner. and ketosis, or those with ischaemic heart disease, One of the important messages from this document is advanced retinal or renal disease. that a regular review of treatment is essential, as a The importance of promoting weight loss in over- deterioration in glucose control over time should be weight patients cannot be overemphasized, since obese expected and this will require an increase in therapy, patients with type 2 diabetes pose several unique with insulin likely to be needed in many patients after problems in diabetic management. First, an increasing a variable period of time after diagnosis. bodyweight makes the attainment of normoglycaemia In 2004, evidence-based clinical guidelines for older by dietary manipulation exceedingly difficult. Second, people were made available [8]. Both levels of ev- both insulin resistance and hyperinsulinaemia will ex- idence and grades of recommendations were given acerbate type 2 diabetes, and may also promote the based on the most up-to-date trial data, although the development of hypertension and dyslipidaemia, which relative lack of major published trials in older people of 16.4 INITIAL TREATMENT 239

GLYCAEMIC TARGET

Frail with high risk of hypoglycaemia and/or Absence of comorbidities: one comorbid chronic condition:

HbA1c: <8.0% HbA1c: <7.0% Fasting glucose: >6.0 ≤8.0 mmol l–1 Fasting glucose: 5.0–7.0 mmol l–1

Receiving lifestyle advice, patient education, engaged in an exercise programme and achieving glycaemic targets?

YES§ NO

BMI <22–25 (underweight + normal BMI >25 overweight weight) Pathway B Pathway A Metformin (except in severe heart failure, Insulin secretagogue (sulphonylurea*, respiratory insufficiency, renal impairment glinides) or metformin depending on (serum creatinine >135 µmol l–1 in men, clinician’s choice or patient characteristics >110 µmol l–1 in women)

Glycaemic targets achieved? Glycaemic targets achieved?

YES §NO YES § NO

Pathway C Pathway C Add insulin • TZD • TZD secretagogue (usually a • α • α-Glucosidase -Glucosidase sulphonylurea) inhibitors inhibitors • • Exenatide Exenatide analogues analogues • DPP-IV • DPP-IV inhibitors inhibitors

Glycaemic targets achieved?

YES § NO

Start insulin

* longer-lasting agents (e.g. chlorpropamide, glyburide (glibenclamide) and sustained-release glipizide) are more likely to cause hypoglycaemia than glipizide, glimepiride and glicazide. § HbA1c monitored every 6 months.

Figure 16.1 Glycaemic targets. antiglycaemic therapies has limited the power of some and heart failure (predominantly attributed to rosigli- of this guidance. A new up-to-date Executive Summary tazone) [20] prompted a reappraisal of their use in of the guidelines is being made available for 2009. A diabetes treatment pathways. Whilst TZDs should not revised algorithm from the European guidelines has be given to patients with a history of serious cardio- been included (Figure 16.1), based on treatment path- vascular disease or heart failure, they are still regarded ways determined by the body mass index (BMI). as very effective glucose-lowering agents and consid- Recent concern with the use of thiazolinediones ered safe when prescribers adhere to guidance. As other (TZDs) in increasing the risk of myocardial infarction recent concerns have included an increased risk of 240 CH 16 EARLY MANAGEMENT OF TYPE 2 DIABETES fractures (mainly in women), the use of TZDs must be often shared by patients until they try insulin. The un- supervised by clinicians with significant diabetes care derlying reasons for the patient’s attitudes include a experience. These agents may be added to treatment horror of injections, awful stories of ‘hypos’, fear of with either metformin or a sulphonylurea (or both). further hospitalization, and the belief that taking insulin Newer agents such as exenatide (an incretin mimetic will change their lives for the worse [22]. It is impera- agent) requires subcutaneous injection, and functions tive that the decision to start insulin be taken after full by activating the glucagon-like peptide 1 (GLP-1) re- discussion with the patient (and carers, as appropriate). ceptor. Amylin analogues such as pramlintide (a syn- Although there are no time limits for when this deci- thetic analogue) offer alternative approaches to glucose sion should be taken, it is suggested that a maximum lowering in elderly patients, although there is limited of 6 months’ perseverance with diet and oral agents be trial evidence of long-term benefit, and they should undertaken before insulin is initiated. In practice, this not be used routinely at this stage. In patients whose decision may have been delayed already for several − BMI is very high (e.g. >35 kg m 2), or in those who years. Able patients can begin insulin at home, much are ‘failing’ on metformin and sulphonylurea therapy, like their younger counterparts, with treatment orga- exenatide may be considered. Inhibitors of the en- nized by a diabetes specialist nurse (whose professional zyme responsible for degrading GLP-1 (dipeptidyl pep- roles are increasing; see Table 16.8), in cooperation tidase IV inhibitors) are also now available (sitagliptin, with the general practitioner. A structured education vildagliptin), and these have consistent effects on low- programme with frequent telephone support during the ering HbA1c with relatively good tolerability and few first 2–3 weeks is essential. Those patients who are un- adverse side effects. Their place in the overall treat- well, or who have other severe medical problems, or ment of older diabetes subjects needs to be determined, where community support is absent, need to be con- however. sidered for hospital admission although this should be A more recent consensus document on the treatment avoided if possible. of type 2 diabetes by the American Diabetes Associa- Usually, treatment can start with about 8–14 units tion and the European Association for the Study of Di- of insulin per day and adjusted thereafter. Although abetes [21] provided an algorithm for insulin initiation NPH insulin is often advised, the recommendation is and recommendations for these newer agents. These to start with a longer-acting analogue such as glargine emphasize a practical goal of lowering the HbA1c to or determir, in combination with an existing oral agent <7%, the effective incorporation of lifestyle modifica- (e.g. metformin). This basal insulin regimen can be tion as an adjunct to management, rapid titration with medications to achieve target glucose levels, and earlier use of insulin where necessary. Table 16.8 Roles of a diabetes specialist nurse for older adults with diabetes. • 16.4.3 Insulin therapy Teaching, advising and counselling patients and carers, both in the clinic and in the patient’s home Few newly diagnosed elderly diabetic subjects require • Educating patients to achieve self-care where possible insulin therapy to sustain life and prevent DKA, al- • though some patients may have a slowly developing Teaching self-monitoring of blood glucose: use of special form of type 1 diabetes and will inevitably require in- techniques for patients with physical disability or visual loss sulin in the future. In everyday clinical practice, the usual indications to start insulin are: • Initiating insulin in the community and instructing patients and informal carers about insulin administration • persisting symptoms with poor patient well-being • Commencement of insulin in the patient’s home • continued weight loss • Liaising with other health professionals to ensure optimal treatment of the patient • a failure to achieve satisfactory glycaemic control with diet and oral agents, usually failing to achieve • Advising residential care home staff about care of residents with diabetes aHbA1c <7.5%. • Providing continuing support and advice to patients and A common error in managing elderly type 2 diabetics carers is an undue reluctance to start insulin therapy – a view 16.5 ESTABLISHING AN INDIVIDUAL DIABETES CARE PLAN 241 tailored to the individual, and should avoid signifi- Urine testing for glucose remains a common prac- cant hypoglycaemia, especially at night. If mealtime tice but is inconvenient, messy, and often misleading glucose levels are a problem, then additional pre-meal because of the raised renal threshold of the elderly. insulin is also required. A number of insulin devices Both, patients and physicians are also often uncertain are available which are easy to use, although in certain about the significance of glycosuria, and the routine cases – such as those patients with confusion, visual use of this parameter is no longer advised. Testing loss or arthritis – the technique of insulin adminis- for the presence of ketones (when poor control is tration should be taught to the spouse or to another present – persistent values of blood glucose >17 mmol relative or friend. l−1, or during severe acute illness) is worth carrying The success of insulin may be evaluated objec- out if patients and informal carers have been suitably tively by factors such as glycaemic control, patient educated regarding its significance. well-being, episodes of hypoglycaemia or the fre- Blood glucose monitoring (e.g. using BM reagent quency of hospital admissions due to diabetes. strip measurements) should be encouraged in all those able to cooperate. Measurements can be taken twice weekly: pre-meal and before-bedtime estimations are 16.5 Establishing an individual ideal, but few patients are this compliant. In other diabetes care plan cases, spouses, district nurses or diabetes specialist nurses may monitor control. The elements of an initial care plan for diabetic elders Guidelines for reasonable diabetic control in the are listed in Table 16.9. This is usually applicable elderly are as follows: a fasting glucose of 5–7 mmol during the first 3–6 months after diagnosis [13]. The l−1, and a random level of 6–8 mmol l−1. These limits care plan should state precisely what the roles of the should allow patients to remain well and be relatively involved individuals are, and where the boundaries free of symptoms of hyperglycaemia, and to avoid the of responsibility lie. The timing and components of risk of hypoglycaemia. It should be remembered that the follow-up can be predetermined, as can the date even glucose levels of 9 mmol l−1 can make some and format of the annual review process, which is a patients feel lethargic, and will need to be lowered. mandatory requirement for all diabetic elders. AHbA1c value <7% should be aimed for. Glycaemic The effective self-monitoring of glycaemic control targets for relatively frail patients with diabetes are is a worthwhile objective for most patients with type provided in Figure 16.1. 2 diabetes, especially for those receiving insulin or who have frequent acute illnesses or hypoglycaemic 16.5.1 Metabolic targeting episodes. In some cases, with the appropriate level of education, patients learn the effects of dietary changes Whilst few clinicians would institute aggressive and exercise on blood glucose levels, by frequent use metabolic control in patients aged >80 years, there is of self-monitoring. increasing evidence of benefits from glucose lowering, blood pressure reduction and lipid lowering in older populations. Therapies for the latter two areas are Table 16.9 Components of an initial diabetes care plan. covered elsewhere in this book. Metabolic targeting 1. Establish realistic glycaemic and blood pressure targets. in geriatric diabetes (see Table 16.10) has a partial 2. Ensure that all parties are agreed on the principal aspects evidence base, and this has been extensively reviewed of diabetes care: patient, spouse or family, GP, informal elsewhere [8]. Most guidelines of diabetes care assume carer, community nurse or hospital specialist, where a single-disease model when recommendations are appropriate. offered, but for older people there is an increased 3. Offer instruction in diabetes self-care management risk of significant other comorbidities and/or frailty, and care must be interpreted on an individual 4. Define the frequency and nature of diabetes follow-up. basis. Patients in these latter categories may be 5. Organize glycaemic monitoring by patient or carer. care-home residents, have evidence of cognitive 6. Refer to social or community services as necessary. impairment, serious mood disturbance, or generally lack self-caring. Unfortunately, about one-third of 7. Provide advice on stopping smoking, exercise, and alcohol intake. patients fall into this latter category, according to the results of a large community-based sample of 242 CH 16 EARLY MANAGEMENT OF TYPE 2 DIABETES

Table 16.10 Metabolic targeting in geriatric diabetes. estimates very carefully in diabetic elders, as it is likely that they will have several other comorbidities that may 1. Independent in self-care, mobile and mentally influence the decision to treat. In addition, applying the alert/single medical disorder: Aim Strict glycaemic, blood pressure and lipid control standard threshold for intervention based on a 10-year risk of coronary heart disease event of 20%, few of 2. Relatively independent with some evidence of functional the older patients with diabetes encountered in every decline and several comorbidities: clinical practice would not require intervention. Aim: Optimize glucose and blood pressure control; consider lowering lipids 3. High dependency and frailty; may be a resident of 16.6 Conclusions a nursing home and/or cognitively impaired; life expectancy <3 years: The management of the older diabetic patient rep- Aim: Symptom control; avoid fatigue; avoid hypogly- resents a major challenge to any physician, whether caemia and intrusive monitoring based in the community or in a hospital setting. Hos- pital physicians without specialist training in diabetes should seek the advice of a consultant diabetologist for people aged >65 years with diabetes, where objective patients whose glycaemic control is persistently unac- measures of dependency were based on the Barthel ceptable or those with severe diabetic complications; ADL score, Extended ADL score and the Mini-Mental for example, extensive foot ulceration, autonomic neu- State Examination score [23]. ropathy or painful neuropathy. Patients with significant diabetic eye disease, such as proliferative or prepro- 16.5.2 Prioritizing diabetes care for liferative retinopathy or maculopathy, require prompt diabetic elders referral to a consultant ophthalmologist. Diabetes care in older adults requires prioritization, A detailed assessment of other cardiovascular risk and a five-step approach is recommended to provide a factors is beyond the scope of this chapter, although framework to develop an individual intervention pro- the presence of hypertension, ischaemic heart disease gramme (Table 16.11). These interventions may in- or hyperlipidaemia may warrant further attention and clude, for example, the aggressive treatment of blood interventions. The development of local specifications glucose and blood pressure, specific rehabilitation pro- for diabetes care, providing an integrated approach to grammes for older people with diabetes, or fast-track management and agreed by all health professionals vascular work-up and early surgical referral [24]. involved, will help this process of referral to occur effi- Charts such as those of Yudkin and Chaturvedi [25] ciently and to provide the most benefit for each patient. permit an estimate of the overall level of vascular risk to be derived which can be used to inform the physician about which thresholds apply for therapeutic interven- References tion. However, it is important to individualize these 1. Sinclair A.J. (ed./Guest ed.) Diabetes in Senior Citi- Table 16.11 Prioritizing diabetes care in older adults: zens: a major threat to personal independence. Br J a five-step approach. Diab Vasc Dis 2005; 5 (1): 3–5. 2. Sinclair AJ and Barnett AH (1993) Special needs of 1. Functional assessment including cognitive testing and elderly diabetic patients. British Medical Journal, 306, screening for depression. 1142–3. 2. Vascular risk assessment with advice on lifestyle modi- 3. Krop JS, Powe NR, Weller WE, Shaffer TJ, Saudek fication and vascular prophylaxis. CD and Anderson GF. (1998) Patterns of expenditures 3. Metabolic targeting (individualized): single-disease and use of services among older adults with diabetes. model versus frailty model. Implications for the transition to capitated managed care. Diabetes Care, 21 (5), 747–52. 4. Consider specific interventions for diabetes-related dis- 4. Holmes J, Gear E, Bottomley J, Gillam S, Murphy M abilities. and Williams R. (2003) Do people with type 2 diabetes 5. Assess suitability for self-care versus carer assistance. and their carers lose income? (T2ARDIS-4). Health Policy 64 (3), 291–6. 16.6 CONCLUSIONS 243

5. Sinclair AJ (1999) Diabetes in the elderly: a perspec- and cardiovascular risk factors in type 2 diabetic tive from the United Kingdom. Clinics in Geriatric patients in rural Costa Rica. Diabetes Care 2003; 26 Medicine, 15, 225–37. (1): 24–9. 6. Hendra TJ and Sinclair AJ (1997) Improving the care 16. Cuff DJ, Meneilly GS, Martin A, et al. Effective of elderly diabetic patients: the ﬁnal report of the St exercise modality to reduce insulin resistance in women Vincent Joint Task Force for Diabetes. Age and Ageing, with type diabetes. Diabetes Care 2003; 26 (11): 26,3–6. 2977–82. 7. Brown AF, Mangione CM, Saliba D and Sarkisian 17. Reaven GM (1988) Role of insulin resistance in human CA. Guidelines for improving the care of the older disease. Diabetes, 37, 1595–607. person with diabetes. J Am Geriatr Soc 2003; 51 (5 18. Asplund K, Wilholm BE and Lithner F (1983) Guidelines): S265–80. Glibenclamide-associated hypoglycaemia: a report of 8. The European Diabetes Working Party for Older Peo- 57 cases. Diabetologia, 24, 412–17. ple. Clinical Guidelines for Type 2 Diabetes Mellitus 19. European Diabetes Policy Group (1999) A Desktop 2001–2004. (Available at: www.eugms.org.) Guide to Type 2 Diabetes Mellitus. International Di- 9. American Diabetes Association. Standards of Medical abetes Federation (European Region), Brussels, Bel- Care 2008. Diabetes Care 2008; 31: S12–54. gium. 10. DECODE Study (Diabetes Epidemiology: Collabora- 20. Singh S, Loke YK and Furberg CD. Long-term tive Diagnostic Criteria in Europe) (1999) Conse- risks of cardiovascular events with rosiglitazone: a quences of the new diagnostic criteria for diabetes in meta-analysis. JAMA 2007; 298 (10): 1189–95. older men and women. Diabetes Care, 22, 1667–71. 21. Nathan DM, Buse JB, Davidson MB, Heine RJ, 11. Apelqvist J. The foot in perspective. Diabetes Metab Holman RR, Sherwin R and Zinman B. Medical Res Rev 2008; 24 (1): S110–15. management of hyperglycaemia in Type 2 diabetes: A 12. Lincoln NB, Radford KA, Game FL and Jeffcoate WJ. consensus algorithm for the initiation and adjustment Evaluation for secondary prevention of foot ulcers in of therapy: A consensus statement of the American people with diabetes: a randomized controlled trial. Diabetes Association and the European Association Diabetologia 2008; 51 (11): 1954–61. for the Study of Diabetes. Diabetes Care 2006; 29 (8): 13. Sinclair AJ, Turnbull CJ and Croxson SCM (1996) 1963–72. Document of care for older people with diabetes. Post- 22. Taylor R (1992) Use of insulin in non-insulin- graduate Medical Journal, 72, 334–8. dependent diabetes. Diabetes Review, 1, 9–11. 14. Knowler WC, Barrett-Connor E, Fowler SE, Hamman 23. Sinclair AJ and Bayer AJ (1998) All Wales Research RF, Lachin JM, Walker EA, Nathan DM and the Dia- in Elderly (AWARE) Diabetes Study. Department of betes Prevention Program Research Group. Reduction Health Report (UK Government), 121/3040, London. in the incidence of type 2 diabetes with lifestyle in- 24. Sinclair AJ (2000) Diabetes in old age: changing con- tervention or metformin. N Engl J Med 2002; 346 (6): cepts in the secondary care arena. Journal of the Royal 393–403. College of Physicians of London, 34, 240–4. 15. Gold-Haber Fiebert JD, Goldhaber-Fiebert SN, Tristan 25. Yudkin JS and Chaturvedi N (1999) Developing risk ML, et al. Randomised controlled community-based stratiﬁcation charts for diabetic and nondiabetic sub- nutrition and exercise intervention improves glycaemia jects. Diabetic Medicine, 16, 219–27. 17 Drug Therapy: Current and Emerging Agents for Hyperglycaemia

Joe M. Chehade and Arshag D. Mooradian Division of Endocrinology Diabetes and Metabolism, University of Florida College of Medicine, Jacksonville, FL, USA

With advanced disease, the achievement of gly- Key messages caemic control can be challenging for both the physician and patient. Often, a combination of two or more • Type 2 diabetes is a heterogeneous disease. oral anti-diabetic agents is needed [5, 6]. Another chal- • The goal for glycaemic control in the elderly lenging aspect of treating type 2 diabetes is targeting diabetic is to achieve euglycaemia without the postprandial hyperglycaemia (PPHG) which, in fact, undue risk of hypoglycaemia. may be a better predictor of diabetes complications • If possible, the management should be tailored than fasting hyperglycaemia. Today, this may be tar- to the individual’s needs. Drug regimens may geted with some newer agents [7]. change over time as the disease progresses. • Postprandial hyperglycaemia is becoming an During the past decade, the pharmacological options important target of the management of diabetes. for treating diabetes have grown considerably such that, today, six classes of agent are available that target hyperglycaemia through different mechanisms 17.1 Introduction (Figure 17.1): Sulphonylureas and meglitinides increase insulin se- The Diabetes Control and Complications Trial cretion from pancreatic β-cells. (DCCT) [1] and the United Kingdom Prospective Biguanides decrease hepatic gluconeogenesis and, to Diabetes Study (UKPDS) [2] have proven the beneﬁt a lesser extent, enhance glucose uptake in skeletal of improved glycaemic control beyond any reasonable muscles. doubt. Furthermore, in elderly people, diabetes control Thiazolidinediones enhance the insulin sensitivity in is an important quality of life issue. Type 2 diabetes is the liver, muscles and adipose tissue. very common in old age, and the rate of development Alpha-glucosidase inhibitors delay carbohydrate ab- of some diabetic complications – including macroan- sorption from the gut. giopathy, nephropathy and neuropathy – appear to be • Incretin mimetics and incretin enhancers accelerated in this age group [3, 4]. stimulate insulin release from the β-cells and

Sulfonylurea Megltinides Incretin-Mimetic DPP-IV inhibitors

Incretin-mimetic Amylin analog INCREASE INSULIN SECRETION DELAY GASTRIC Biguanides EMPTYING

ENHANCE DECREASE INSULIN HEPATIC SENSITIVITY GLUCOGENESIS DIABETES CONTROL Thiazolidinediones

REDUCE FOOD INTAKE & PROMOTE DELAY CHO Alpha-glucosidase SATIATY ABSORPTION inhibitors

Incretin-mimetic DECREASE Amylin analog POST-PRANDIAL GLUCAGON SECRETION

Incretin-mimetic DPP-IV inhibitors Amylin analog

Figure 17.1 Mechanisms of action of the currently available oral agents for the treatment of type 2 diabetes. CHO = carbohydrate.

inhibit glucagon secretions from alpha cells in a and chlorpropamide. These so-called ‘ﬁrst-generation’ glucose-dependent manner. sulphonylureas were followed by the development of • Pramlintide, by modulating gastric emptying, can ‘second-generation’ sulphonylureas such as glyburide prevent the postprandial rise in glucagon and en- or glibenclamide, glipizide, gliclazide, gliquidone and hance satiety. glimeperide (Table 17.1). Although additional pharmacological agents are re- The primary mechanism of action of sulphonylureas quired to manage other risk factors such as hyper- is through the depolarization of pancreatic β-cells by tension and dyslipidaemia, in this chapter attention is blocking ATP-dependent potassium channels, causing focused only on those agents that control blood glu- an inﬂux of calcium and stimulating insulin release [8]. cose. A decrease in plasma glucagon levels was also noted in some studies [9]. Although other extra-pancreatic effects have also been described, sulphonylureas are 17.2 Sulphonylureas ineffective in patients with type 1 diabetes who lack islet β-cells [10]. The reported increased insulin sen- Sulphonylureas have been used extensively world- sitivity in peripheral tissues is thought to be mainly wide since the introduction of tolbutamide and carbu- a consequence of the reduction of hyperglycaemia and tamide in 1956. Shortly thereafter, other compounds an alleviation of glucotoxicity [11, 12]. A reduced hep- were developed, including acetohexamide, tolazamide atic insulin extraction was demonstrated with glipizide 17.2 SULPHONYLUREAS 247

Table 17.1 Comparative pharmacological proﬁle of sulphonylurea agents. Drug Dose range Duration of Doses Metabolism (mg) action (h) per day First-generation Acetohexamide 250–1500 8–24 2 60% hepatic, active renal metabolism Tolbutamide 500–3000 6–12 2–3 Hepatic with renal excretion Tolazamide 100–1000 12–24 1–2 Hepatic with renal excretion Chlorpropamide 100–500 24–72 1 30% renal excretion, some hepatic metabolism Second-generation Glipizide 2.5–40 16–24 1–2 Hepatic, renal excretion of inactive metabolites Glipizide XL (Glucotrol 5–20 24 1 Hepatic, renal excretion of inactive metabolites XL) Glyburide 1.25–20 12–24 1–2 Hepatic with renal excretion of active metabolites Glyburide Micronized 1.25–10 12–24 1 Hepatic with renal excretion of active metabolites (Glynase) Glimepiride 1–8 12–24 1 Hepatic with renal excretion of active metabolites Gliquidone 15–60 8–10 1–2 Hepatic with renal excretion of inactive metabolites Gliclazide 40–320 10–15 1–2 Hepatic with renal excretion of inactive metabolites

[13] and glibenclamide [14]. However, this effect is chlorpropamide. While the long-acting preparation of probably secondary to increased insulin secretion. glipizide (glipizide GITS) can be administered once The pharmacological characteristics of the sulphony- daily, its longer duration of activity may result in a lureas are summarized in Table 17.1. They are rapidly higher incidence of hypoglycaemia. absorbed, and detectable levels can be found in the On average, sulphonylureas lower plasma glucose serum within an hour of ingestion. They are 90–99% by 3.3–3.9 mmol l−1 (60–70 mg dl−1), with a concomi- bound to serum proteins, mainly albumin [15], and can tant reduction in HbA1c of 1.5–2% [8,17]. Some be displaced by other drugs. On a milligram per mil- 10–15% of subjects with type 2 diabetes fail to re- ligram basis, second-generation sulphonylureas have spond to this class of agent – a situation known as 100-fold the potency of the first-generation agents [16] ‘primary failure’. Annually, an additional 5–10% of (Table 17.1). Moreover, since their binding to plasma initial responders have ‘secondary failure’ of response proteins is mainly to non-ionic sites, drug–drug inter- to sulphonylureas [18]. The most frequent side effect actions are less likely than with first-generation agents. of sulphonylureas is hypoglycaemia [19]. UKPDS data Even though there are some differences in the pharma- showed that, over a six-year treatment period with cokinetic and pharmacodynamic properties of various sulphonylureas, 45% of patients experienced at least sulphonylureas, to date no clinical evidence has been one hypoglycaemic episode, and 3% had severe hy- submitted of the superiority of any particular agent. poglycaemia [20]. It is noteworthy that continuous Sulphonylureas are mainly converted by the liver to dextrose infusions for several days may be required inactive, less active – or, in the case of acetohexamide, to treat sulphonylurea-induced hypoglycaemia. A ret- more active – metabolites. First-generation drugs and rospective study conducted by the Swedish Board of their metabolites are largely excreted by the kidney, Health and Welfare between 1975 and 1982 revealed while second-generation agents are excreted via both that all serious causes of hypoglycaemia secondary the kidney and bile in varying proportions (Table 17.1). to sulphonylureas were associated with advanced age, Although differences in absorption, metabolism and drug–drug interactions and acute energy deprivation elimination do not affect long-term efficacy, they are due to gastroenteritis [21]. Due to the high prevalence important in terms of the frequency and severity of ad- of polypharmacy in elderly people, caution should be verse effects in elderly people. Of special concern is the taken when sulphonylureas are combined with agents prolonged half-life of chlorpropamide, along with its that potentate their efficacy. water-retaining properties. Of note is that the UKPDS Tolazamide and glyburide have active metabolites showed that the incidence of hypoglycaemia with gly- that accumulate when creatinine clearance is reduced. buride (glibenclamide) therapy was as high as with Low-dose tolbutamide and glipizide may therefore be 248 CH 17 DRUG THERAPY: CURRENT AND EMERGING AGENTS FOR HYPERGLYCAEMIA preferable for elderly patients with mild renal insuffi- of decreased renal function. No pharmacokinetic ciency. However, because of its short half-life, tolbu- studies have been conducted in subjects with hepatic tamide may need to be given three times a day, but insufficiency. compliance with such regimen is often difficult. None Unlike sulphonylureas, metformin has no insulino- of the sulphonylureas should be used in patients in tropic effect and does not cause hypoglycaemia when −1 whom the creatinine clearance is less than 30 ml min used as monotherapy [28]. As such, it is considered [22]. as an antihyperglycaemic agent rather than a hypo- Other potential disadvantages of sulphonylureas are glycaemic agent. The antihyperglycaemic action of the associated hyperinsulinaemic state and weight metformin is primarily due to a reduction in hep- gain. Over a nine-year period, subjects assigned to atic gluconeogenesis, thereby reducing basal glucose sulphonylureas in the UKPDS gained on average output. To a lesser extent, it enhances glucose up- 4 kg compared to 2 kg for those treated with diet take in the peripheral tissue, mainly muscle [25, 28]. alone. In contrast, people in the metformin arm main- The reduction of fatty acid oxidation in metformin- tained a steady weight [2, 20]. This side effect of treated subjects may account for some of its antihy- sulphonylureas is undesirable in obese subjects, bit in perglycaemic action [29]. older subjects – where being underweight is common A dose–response relationship of the antihypergly- [23] – the weight gain-promoting properties of sulphonylureas may not be a serious deterrent to their use. caemic effect of metformin was demonstrated in a However, obesity [body mass index (BMI) >30gm−2] double-blind, placebo-controlled trial over 14-week continues to increase mortality risk at any age [24]. The period. At doses between 500 to 2000 mg daily, potential advantages of sulphonylureas include their metformin reduced the adjusted mean fasting plasma −1 long track record, ease of use, and affordability. How- glucose (FPG) from baseline by 1.04–4.62 mmol l −1 ever, hypoglycaemia remains the main limiting factor (19–84 mg dl ) and adjusted mean HbA1C by in elderly people. 0.6–2.0% [30]. The maximum efficacy is observed with a daily dose of 2000 mg [30]. Metformin has been used extensively both as 17.3 Biguanides monotherapy and in combination with sulphonylureas or insulin [31, 32]. It is noteworthy that individuals Metformin (dimethylbiguanide) and phenformin with primary or secondary failure of sulphonylureas (phenethyl biguanide) were introduced for the are unlikely to respond to metformin alone. However, treatment of type 2 diabetes in 1957, and buformin when metformin is combined with sulphonylureas in was introduced a year later [25]. During the 1970s, individuals who appear to have secondary failure, a phenformin and buformin were withdrawn in many substantial blood glucose lowering occurs [33]. countries due to their association with lactic acidosis Metformin has a favourable effect on plasma lipid [26]. Metformin is now the biguanide of choice and levels and on blood pressure [34]. There is a moderate is extensively used worldwide. The bioavailability reduction in triglyceride levels due to decreased of metformin after a standard dose of 500 mg hepatic synthesis of very low-density lipoprotein is approximately 50–60%. The presence of food (VLDL)-cholesterol and a slight increase in the decreases the extent and slightly delays absorption high-density lipoprotein (HDL) level [35]. Perhaps of the drug. In contrast to sulphonylureas, metformin is not significantly bound to plasma proteins, and the most attractive feature of this agent is that it is steady-state plasma concentrations are achieved often associated with weight loss, and so counteracts within 24–48 h of dosing [27]. Metformin does not the weight gain associated with sulphonylureas or undergo any hepatic metabolism and is excreted insulin therapy [31]. However, obesity is much unchanged in the urine [25]. Because the renal less of a problem in elderly people than in the clearance of metformin is 3.5-fold greater than the middle-aged [24, 36]. In the UKPDS, metformin creatinine clearance, tubular secretion appears to be monotherapy in overweight subjects was the only arm the main route of elimination. In healthy elderly of the study which showed a significant reduction subjects, the total plasma clearance is decreased and in cardiovascular events and overall mortality [31]. thus the half-life of the drug is prolonged. These The precise underlying mechanism of this favourable pharmacokinetic changes with ageing are the result outcome is not known, however. 17.4 ALPHA-GLUCOSIDASE INHIBITORS 249 The most common and troublesome side effects acarbose and miglitol are now available in many coun- of metformin include gastrointestinal discomfort, nau- tries; a third agent, voglibose, is available in Japan. sea, diarrhoea, anorexia, and rarely, a metallic taste. Alpha-glucosidases are hydrolase enzymes within Starting therapy with 500 mg daily and increasing the brush border of the small bowel, and are re- the dose gradually can attenuate these side effects. sponsible for cleavage of the non-absorbable oligo- The biguanide-associated malabsorption of vitamin B12 and dissaccharides into monosaccharides, which are (cyanocobalamin) and folate is usually not of a major then rapidly absorbed from the gastrointestinal tract. clinical concern [37]. However, this should be borne in The improvement of postprandial hyperglycaemia is mind when prescribing for elderly subjects who have through a reversible inhibition of the brush border a relatively high incidence of atrophic gastritis and vi- glucosidases, resulting in a redistribution of carbohydrate absorption from the upper portion of the gut to a tamin B12 deficiency. Although rare, the most dreaded side effect is lactic acidosis, the incidence of which more extended surface area covering the whole length is approximately 9 per 100 000 persons per year in of the small intestine [40]. However, this may result metformin users [38], almost 10-fold lower than that in a higher level of fermentable carbohydrate reach- associated with phenformin. Therefore, any clinical ing the large bowel where they are metabolized by condition associated with, or predisposing, to lactate colonic microflora to short-chain fatty acids and then generation or a decreased ability to clear lactate is a absorbed [41]. There is no substantial caloric loss in the faeces. Due to the high specificity of these agents contraindication to the use of metformin. These con- for α-glucosidases, β-glucosidases, like lactoses, are ditions include hepatic disease, alcoholism, congestive not inhibited and lactose intolerance is not a clinical heart failure, peripheral vascular disease, obstructive problem. Based on its mode of action, several infer- airway disease and particularly renal impairment, all of ences can be made. One is that these agents should which are relatively common in elderly people. Met- be effective in every individual who has postprandial formin is absolutely contraindicated when the serum hyperglycaemia and ingests sufficient amounts of car- creatinine level is >132 mmol l−1 (1.5 mg dl−1) in men −1 −1 bohydrates. The second inference is that this agent will and >124 mmol l (1.4 mg dl ) in women. Because be effective only when given with meals. Being a com- the risk of lactic acidosis increases with the degree petitive inhibitor of carbohydrate digestion, it will not of renal impairment and the patient’s age, metformin be effective when given on an empty stomach. The should not be initiated in patients aged over 80 years. third inference is that AGIs may cause gastrointestinal Above the age of 70, careful monitoring of renal func- flatulence and loose stools in some individuals. Less tion and at least a baseline creatinine clearance is a than 2% of an oral dose of acarbose is absorbed as ac- requirement. Metformin should be promptly withdrawn tive drug. Acarbose is metabolized exclusively within with the acute onset of dehydration, hypoxaemia, sep- the gastrointestinal tract, primarily by the intestinal mi- sis or the use of contrast media. In this setting, met- croflora and to a lesser extent by digestive enzymes. A formin should not be reinstituted unless renal function fraction of these metabolites (34% of the oral dose) is is demonstrably normal. absorbed and subsequently excreted in the urine [42]. Although the efficacy of metformin is comparable Other mechanisms of action of acarbose may include to that of sulphonylureas and its use is not limited an enhancement of glucagon-like peptide-1 (GLP-1) by hypoglycaemia, for the above reasons it is not response; however, this effect is probably a minor necessarily the agent of first choice for many older contributor (if any) to the antihyperglycaemic effects patients with type 2 diabetes. of these agents [43]. Miglitol absorption is saturable at high dose. When an oral dose of 25 mg is given, it is almost com- 17.4 Alpha-glucosidase inhibitors pletely absorbed, whereas a dose of 100 mg is only 50–70% absorbed. Miglitol is not metabolized and is More than 50% of type 2 diabetic patients receiving eliminated unchanged by renal excretion. The elim- oral conventional therapy have persistent postprandial ination half-life is 2 h. In renal failure, the miglitol hyperglycaemia [39]. The alpha-glucosidase inhibitors plasma concentration increases but a dose adjustment (AGIs) are a class of agent that primarily target post- is not recommended because the drug acts locally. prandial hyperglycaemia. Acarbose was the first such Plasma concentrations of both acarbose and miglitol agent to be made available commercially, and both are increased in renal impairment, and no safety data 250 CH 17 DRUG THERAPY: CURRENT AND EMERGING AGENTS FOR HYPERGLYCAEMIA are available when creatinine clearance is less than 17.5 Thiazolidinediones 25 ml min−1 [44]. Acarbose treatment reduced HbA1C by 0.5–0.7% Thiazolidinediones (TZDs) enhances insulin sensitiv- in monotherapy compared to 0.5–1.2% when com- ity in the liver, adipose tissue and muscle, with- bined with a sulphonylurea or biguanide [45, 46]. out affecting insulin secretion. Their site of action is On average, the postprandial glucose surge was re- mediated through selective activation of peroxisome −1 −1 duced by 2.7–3.3 mmol l (50–60 mg dl ), but the proliferator-activated receptor gamma (PPAR-γ), a nu- effect on fasting glucose levels was more modest at clear receptor that plays an important role in adipoge- −1 −1 1.1–1.6 mmol l (20–30 mg dl ) [45]. nesis [53]. So far, the precise mechanisms that lead to AGIs are best tolerated when started at a low dose the transcriptional regulation of genes involved in in- (25 mg once daily with the beginning of the meal), and sulin sensitization and lipid metabolism have not been then increased gradually over a 6- to 8-week period to a elucidated. maximum dose of 100 mg t.i.d. with meals. In subjects Troglitazone, the first compound in this class to be- weighing <60 kg, the total daily dose should not ex- come available, was introduced in 1997 but withdrawn ceed 150 mg. In older subjects, a dose–response study from the market shortly afterwards due to severe hep- showed that the efficacy of acarbose is near-maximal atotoxicity. Two other agents, rosiglitazone and piogli- at 25 mg when the meal size does not exceed 483 kcal tazone, are still available. The TZDs are antihypergly- and contains only 61 g of carbohydrates [47]. Similarly, caemic agents, and do not cause hypoglycaemia when in an elderly population with a median age of 70 years, used as monotherapy. However, TZD monotherapy miglitol achieved near-maximal metabolic benefits at a will also fail in patients without enough endogenous or low dosage of 25 mg t.i.d. [48]. The Precose Resolu- exogenous insulin. The TZDs are 99% bound to plasma tion of Optimal Titration to Enhance Current Therapies proteins, notably albumin, while the plasma elimina- (PROTECT) study showed that there was no difference tion half-lives range between 3 and 7 h for rosiglita- in efficacy or safety of acarbose when subjects aged zone and pioglitazone. In the case of rosiglitazone, a over 60 years were compared to younger subjects [49, twice-daily regimen (4 mg b.i.d.) was more efficacious 50]. In a study of patients with impaired glucose toler- than a once-daily dosing (8 mg q.d.), with reductions in ance (IGT), acarbose treatment was associated with a HbA1c from baseline of 0.7% and 0.3%, respectively significant reduction in the risk of cardiovascular dis- [54]. ease and hypertension [51]. All TZDs undergo major hepatic metabolism. Unlike Like bigaunides, this class of agents is considered to troglitazone, rosiglitazone does not appear to induce be antihyperglycaemic, since when used as monother- cytochrome P450 (CYP)3A4-mediated metabolism and apy, they do not result in hypoglycaemia. Another is, therefore, likely to experience fewer drug–drug in- potential advantage of AGIs when used as monother- teractions. The TZDs are mainly excreted in bile, either apy is that there is no associated hyperinsulinaemia unchanged or as metabolites, and then eliminated via or weight gain. However, if patients are treated with the faeces. Consequently, there is no need for dose a combination of an AGI and a hypoglycaemic agent adjustment in patients with renal impairment. On the such as insulin or a sulphonylurea, glucose should be other hand, the TZDs are contraindicated in patients used to treat the hypoglycaemic reactions since sucrose with active liver disease or if serum aminotransferase or a complex carbohydrate will not be readily effective. (ALT) is more than 2.5-fold the normal level at base- Hypoglycaemia is a major concern and a limiting line. factor in treating elderly patients with type 2 diabetes. The efficacy of the TZDs has been demonstrated Considering the favourable tolerability and safety proin multiple clinical trials [55, 53]. In monotherapy file of AGIs, some diabetologists therefore choose with troglitazone or rosiglitazone, the mean decline these agents as first-line therapy in elderly type 2 di- in HbA1C from baseline ranged between 0.6 and 1% abetic subjects with fasting blood glucose levels of [55]. A more pronounced drop in HbA1C of 1.0–1.4% <11.0 mmol l−1 [52]. However, the relatively higher was noted when combined with insulin, however. A cost of these agents, the need for multiple daily dos- head-to-head comparison study for pioglitazone and ing and the adverse gastrointestinal side effects have rosiglitazone has now been conducted [56] wherein the limited their widespread use. mean decline in HbA1C from baseline at 24 weeks was 17.6 MEGLITINIDES 251 0.7% and 0.6% for pioglitazone and rosiglitazone, re- acknowledged the increase risk of cardiac events asso- spectively. Both LDL- and HDL-cholesterol tend to ciated with rosiglitazone, especially in a subgroup of increase with TZD therapy, while serum triglyceride patients, yet opted to keep the drug on the US market levels may decrease with troglitazone and pioglitazone with the addition of a ‘black box’ warning. but not with rosiglitazone. Although rosiglitazone has a The TZDs are usually well tolerated and, when used pure PPAR-γ agonist activity, pioglitazone and trogli- as monotherapy, there is no associated risk of hypo- tazone have in addition some PPAR-α activity that may glycaemia. Peripheral oedema is occasionally seen in contribute to their triglyceride-lowering effect. Com- monotherapy, more so with the concomitant use of pared to baseline, the reduction in triglyceride level insulin, mainly as a result of the plasma volume ex- may be as high as 26% with troglitazone and piogli- pansion. Therefore, these agents are not recommended tazone therapy [53]. The effect of pioglitazone and in patients with New York Heart Association (NYHA) rosiglitazone is different on plasma lipids, independent Class III or IV congestive heart failure. A minor de- of glycaemic control or concomitant lipid-lowering or crease in haematocrit and haemoglobin is another ob- other antihyperglycaemic therapy. Pioglitazone, com- servation correlates with the dilutional effect of fluid pared with rosiglitazone, is associated with significant retention. improvements in triglycerides, HDL-cholesterol, LDL The once-daily dosing regimen and lack of asso- particle concentration and LDL particle size [56]. ciated hypoglycaemia make the TZDs an attractive On average, 20–50% of individuals fail to respond option in elderly subjects. On the other hand, recent to this class of agent [55]. It appears that obese and concerns about cardiovascular safety, suboptimal effi- hyperinsulinaemic individuals are more likely to re- cacy in underweight subjects and costs may limit their spond than those who have insufficient insulin. When usefulness, particularly as a first-line agent. combined with sulphonylureas, TZDs were as effective as the combination of metformin and sulphonylureas [57, 58]. A major disadvantage of the combination of 17.6 Meglitinides TZDs and sulphonylureas is the associated weight gain, which may be as high as 6.5 kg after one year of ther- Meglitinide belongs to a novel group of insulinotropic apy. A modest increase in body weight is also observed agents known as the non-sulphonylurea insulin when TZDs are used as monotherapy. secretagogues [63]. Both, repaglinide and mitiglinide The major concern of serious hepatotoxicity seen are benzoic acid derivatives, while nateglinide is a with troglitazone seems not to be a class-related issue. phenylalanine derivative. As with the sulphonylureas, After eight years of post-marketing, the available data the insulinotropic effect is mediated through the suggest that pioglitazone and rosiglitazone do not have ATP-regulated potassium channels but via different an increased incidence of hepatotoxicity. During a binding sites on the β cells [64]. After oral admin- 48-week study with rosiglitazone and pioglitazone, istration, meglitinides are rapidly absorbed from the improving insulin sensitivity resulted in improved the gastrointestinal tract, with maximum plasma histological markers of non-alcoholic steatohepatitis concentrations being reached within 0.8 h and the drug (NASH) [59, 60]. rapidly eliminated with an approximate half-life (t ) 1/2 The controversy surrounding the increased risk of of 1–1.7 h [65, 66]. However, the wide variability cardiovascular events with rosiglitazone has raised ma- of t of repaglinide elimination kinetics ranging 1/2 jor safety concerns, and caused a major dilemma for between 0.5 and 8 h is of concern. The absorption practitioners. In a meta-analysis study, rosiglitazone of nateglinide is increased when the medication is was associated with a significant increase in the risk of taken 10 min before a meal, whereas the Cmax is myocardial infarction, with an odds ratio of 1.43 (95% decreased by 34% and the Tmax increased by 22% CI: 1.03–1.98; P = 0.03) [61]. In a prospective ran- when the drug is given after the meal. The absorption domized trial of 5238 patients with type 2 diabetes is also reduced during fasting. The influence of food and macrovascular disease, pioglitazone did not in- on the pharmacokinetics of repaglinide is of lesser crease the risk of myocardial ischaemia. Rather, there importance. was a trend towards a benefit in regard to all-cause The pharmacokinetic profile makes meglitinides a mortality, non-fatal myocardial infarction, stroke and suitable agent for targeting postprandial hypergly- acute coronary syndrome [62]. On 30th July 2007, the caemia [7]. Meglitinides are completely metabolized United States Food and Drug Administration (FDA) by the liver, with 90% of repeglinide being excreted 252 CH 17 DRUG THERAPY: CURRENT AND EMERGING AGENTS FOR HYPERGLYCAEMIA via the biliary route and only 8% via the urine. In con- to a small dose of short-acting sulphonylurea is still trast, 90% of nateglinide is excreted via the urine and not clear [75]. 10% via the bile [65, 66]. When comparing healthy young individuals with people aged ≥65 years, there was no difference in the pharmacokinetic parameters. 17.7 Incretin mimetics and However, elderly type 2 diabetic subjects had a sig- enhancers: Glucagon-like nificantly higher mean diurnal plasma concentration peptide-1 and dipeptidyl and a reduced clearance when compared to healthy peptidase IV inhibitors controls [67]. Although the area under the plasma concentration–time curve (AUC) and C were sig- max Over 40 years ago, the observation that enteral glu- nificantly increased in subjects with various degrees cose provided a more pronounced insulinotropic effect of renal impairment, a dose adjustment was not nec- compared with an isoglycaemic intravenous challenge essary. However, the authors recommended a careful led to the development of the incretin concept [76]. and gradual increase in the initial dose. Following a Incretins are intestinal hormones which are released single dose of repaglinide, patients with moderate to after meal ingestion and play an important role in severe liver disease had higher and more prolonged normal glucose homeostasis. Glucose-dependent in- serum concentrations [68]. Although liver disease did sulinotropic polypeptide (GIP) – the first incretin to not significantly increase the risk of hypoglycaemic be identified – is synthesized in duodenal and jejunal episodes in this latter study, the safety of repaglinide enteroendocrine K cells in the proximal small bowel. remains questionable in this subgroup of patients. Glucagon-Like Peptide-1 (GLP-1) is a very potent in- A 1-year comparison study of repaglinide and glimepiride, with an 8-week titration period, showed sulinotropic peptide hormone secreted by the L-cells of the intestinal mucosa in the lower gut [77, 78]. similar reductions of HbA1c (–1.2% versus – 1.1%) and FPG levels (–2.11 versus – 2.7 mmol l−1), but The insulinotropic effect is mediated at the level of repaglinide resulted in a greater reduction in FPG level the L-cells through a stimulation of adenylate cyclase than glimepiride (–2.56 versus -1.17 mmol l−1) [69]. and protein kinase A activity [79]. Unlike sulphony- When the efficacy of repaglinide was compared to lurea agents, GLP-1 has no hypoglycaemic effect in glyburide in a multicentre, randomized, double-blind the absence of glucose [80]. Its secretion increases in response to unabsorbed nutrient within the intestinal lu- study over one year, the reduction in HbA1C was similar in both groups [70]. However, when added men. In addition, it has an inhibitory effect on glucagon to metformin in suboptimally controlled diabetes secretion and the gastric emptying rate [77]. Although fasting hyperglycaemia can be reduced with GLP-1, its (HbA1C = 8.5%), the HbA1C fell by 1.4% [71]. A combination therapy achieved a better control primary target is postprandial hyperglycaemia [81, 82]. than either drug alone. Repaglinide and nateglinide In clinical studies, the two shorter forms of GLP-1, have been also tested in association with TZD the (7–37) and (7–36) amides, have been used. These [72, 73]. As with the sulphonylureas, weight gain peptides have a short duration of action (<1min)and and hypoglycaemia were the two most frequent should be given parenterally. Although, both GLP-1 adverse effects, although there was a reduced risk and GIP act as incretin hormones in healthy subjects, of hypoglycaemia when a meal was omitted and the more attention has been focused on GLP-1 because repaglinide dose withheld. [74]. of the observation that in type 2 diabetes the GLP-1 The potential for a reduced risk of hypoglycaemia is level is reduced yet the response to GLP-1 is preserved, an interesting feature of this class of agent, especially whereas GIP secretion is normal but the response to in older individuals. However, additional studies are re- GIP is blunted or absent. Circulating levels of both quired to define the relative value of these agents. Of GLP-1 and GIP decrease rapidly due to enzymatic concern is the wide range of variability in drug elimi- degradation, mainly by the action of dipeptidyl pep- nation kinetics with repeglinide. In addition, a prepran- tidase IV (DPP-IV) and renal clearance. dial dosing regimen may be an obstacle to achieving Exenatide is the synthetic version of exendin-4, an long-term compliance in some individuals. Nateglinide incretin hormone originally found in the saliva of the pharmacokinetics are more favourable in terms of im- Gila monster, Heloderma suspectum. Exenatide has a proving postprandial hyperglycaemia. Finally, the real greater potency and a longer duration of action than the advantage of these relatively costly agents compared native GLP-1 when administered subcutaneously [83]. 17.8 AMYLIN ANALOGUES 253 Exenatide shares with the native GLP-1 its biological (GI) side effects were infrequent and hypoglycaemic actions and is resistant to DPP-IV degradation. Follow- episodes very rare and insignificant [91, 92]. The first ing subcutaneous (sc) administration, it reaches median DPP-IV inhibitor (sitagliptin) was approved for the peak plasma concentrations in 2.1 h, and is eliminated treatment of type 2 diabetes in the United States in predominantly by glomerular filtration. No dose adjust- October 2006. ment is required in mild to moderate renal impairment. Further clinical experience is required before the Exenatide is administered sc within a 60-min period incretin approach can be accepted as an alternative before the two main meals of the day, 6 h or more to the currently available hypoglycaemic agents to apart, at a starting dose of 5 μg twice daily (b.i.d.) and treat the elderly population with type 2 diabetes. Al- then titrated to 10 μg b.i.d. after 4 weeks [84]. The though the risk of hypoglycaemia with GLP-1 ana- main adverse events are nausea, vomiting, diarrhoea, logues is less compared to sulphonylureas and it has headache and dizziness. The risk of hypoglycaemia is a favourable weight change profile, the sc route of increased when exenatide is used in combination with administration and the frequent GI side effects may sulphonylureas. The drug was approved by the FDA be problematic. For DPP-IV inhibitors, the once-daily in April 2005 as an adjunct therapy for type 2 diabetes regimen, the lack of weight gain and almost no hy- in conjunction with sulphonylureas, biguanides and poglycaemia may make this class of agents an ex- TZDs. At 30 weeks, the HbA1c reduction from base- cellent first-line therapy, especially in those patients line was between -0.8% and -0.9% in the metformin, with mild glucose elevation. However, cost reduc- sulphonylurea and metformin-sulphonylurea combination, careful post-marketing surveillance for adverse tion subgroup, with weight reductions of 1.6 kg in the effects and continued evaluation in longer-term stud- sulphonylurea and metformin-sulphonylurea arms, and ies are required to determine the role of this new class 2.8 kg in the metformin-alone arm [85, 86, 87]. The among current pharmacotherapies. The recommended reductions in HbA1c and body weight were sustained dose of sitagliptin is 100 mg once daily, with or with- at 2.5 years at -1% and 5 kg, respectively [88]. In com- out food, as monotherapy or as combination therapy bination with TZD, the HbA1c reduction was -0.8% at with metformin or a TZD. A dosage adjustment is rec- 16 weeks with a 1.5 kg weight loss [88]. The efficacy ommended in patients with moderate or severe renal of a long-acting release (LAR) exenatide is currently insufficiency or with end-stage renal disease (ESRD) undergoing Phase III studies at a weekly dose of 0.8 requiring dialysis. In these patients, the initial dose to 2 mg. should be 50 mg once daily if the creatinine clearance Due to the rapid degradation of GLP-1 by DPP-IV, (CrCl) is ≥30–<50 ml min−1, and 25 mg once daily if attempts have been made to develop specific in- <30 ml min−1. No dosage adjustment is required based hibitors. DPP-IV is a ubiquitous membrane-spanning, on age; however, because sitagliptin is excreted sub- cell-surface aminopeptidase that is expressed in many stantially by the kidney it may be useful to assess renal tissues, such as liver, lung, kidney, lymphocytes, en- function in elderly patients prior to initiation, and pe- dothelial cells and intestinal brush border [89]. DPP-IV riodically thereafter. catalyzes the cleavage of GIP and GLP-1 to bioinac- tive GIP3−42 and GLP−19–37 or GLP−19−36 amide, respectively [90]. Two DPP-IV inhibitors, vildagliptin 17.8 Amylin analogues and sitagliptin, have completed Phase III clinical trials, and their efficacy in both monotherapy and combina- Amylin was identified in 1987 [93] as a 37-amino acid tion therapy regimens was examined. Both, vildagliptin hormone secreted in conjunction with insulin by the and sitagliptin significantly lowered HbA1c when used pancreatic β-cells in response to a glucose load or other either as an initial monotherapy, or in combination with insulin secretagogues [94]. In type 1 diabetes, the re- other oral anti-diabetic agents such as biguanides, thi- duction in amylin concentration parallels the decline azolidinediones or sulphonylureas. The average fall in in insulin secretion [95]. When compared to a popu- HbA1c in most reports was between 0.5 and 0.8% from lation of lean, healthy subjects, basal and stimulated baseline, with no associated weight gain. DPP-IV in- amylin secretion were significantly higher in obese pa- hibitors had an increased risk of infection [risk ratio 1.2 tients with or without impaired glucose tolerance [96]. (95% CI: 1.0–1.4) for nasopharyngitis and 1.5 (95% However, in long-standing type 2 diabetes, the amylin CI: 1.0–2.2) for urinary tract infection] and headache concentration decreases in conjunction with β-cell fail- [risk ratio 1.4 (95% CI: 1.1–1.7)]. Gastrointestinal ure [97]. The effect of amylin on glucose metabolism 254 CH 17 DRUG THERAPY: CURRENT AND EMERGING AGENTS FOR HYPERGLYCAEMIA tors; -cells, -cells headaches and upper respiratory infections adjustment with renal impairment release from β Decrease post-prandial glucagon secretion from α Nasopharyngitis, inhibitors Need dose Stimulate insulin 0 neutral 1 − ↓↓ –cells α 30 ml min -cells, Decrease s; AGI: Alpha-glucosidase inhibi (nausea, vomiting and diarrhoea) release from β post-prandial glucagon secretion from Creatinine clearance < Inhibits gastric emptying Reduces food intake Gastrointestinal, Stimulate insulin + Flatulence inhibition of the intestinal brush border glucosidase or neutral neutral to Gastrointestinal Reversible Diarrhoea Gastroparesis artery disease; TZD: thiazolididione 1.5 0.5–0.7 0.8–0.9 0.6–0.7 increase risk of CHF, may increase risk of CAD (Rosiglita- zone) sensitivity in muscles and liver Liver failure Enhance insulin NYHA III-IV ↑↑ ↓ gluconeogenesis impairment Liver failure Hypoxaemic states (risk of lactic acidosis) or neutral Decrease hepatic Renal 000 -cells insulin release from β (risk of hypoglycaemia) Hypoglycaemia Gastrointestinal Oedema, slight Stimulate Renal failure -cells poglycaemia insulin release from β (risk of prolonged hypoglycaemia) ↑↑ ↑↑ ↓ Sulphonylureas Meglitinides Biguanides TZD AGI1.5–1.8 1.0–1.5 GLP-1 1.0–1.5 DPP-IV 0.7– + ++ + +++ ++ +++ +++ +++ ++ Comparative profile of available oral agents for treatment of type 2 diabetes (in monotherapy). in ↓ 1c change HbA CHF: Chronic heart failure; NYHA: New York Heart Association; CAD: Coronary Side effects Prolonged hy- GLP-1: Glucagon-like-peptide-1; DPP-IV: Dipeptidyl peptidase IV. Table 17.2 Mode of action Stimulate Contraindications Renal failure Body weight Hypoglycaemia Cost Potency 17.9 INSULIN AND INSULIN ANALOGUES 255 involves several mechanisms. By slowing the rate of (P < 0.027) or q.i.d. (P < 0.040) treatment groups, gastric emptying, it limits the proportion of nutrients compared to a 0.8 kg weight gain in the placebo group delivered to the gut, thus preventing the postprandial [103]. hyperglycaemic surge [98]. Amylin also suppresses The frequency and the route of administration of postprandial glucagon secretion, helps replenish glyco- pramlintide have raised major concerns regarding the gen stores [7, 99], promotes satiety, and reduces caloric suitability of such agents in the elderly population. intake. However, due to its short half-life and its tendency to aggregation, amylin is not suitable for clinical use. 17.9 Insulin and insulin analogues The amylin analogue ‘pramlintide’ is an injectable synthetic analogue that is structurally identical to hu- As insulin therapy in diabetes is discussed extensively man amylin, but with the exception of a proline substi- in Chapter 18, this section will deal mainly with the tution at positions 25, 28 and 29 [100]. Clinical studies role of some of the newer insulin preparations. Multiple have demonstrated a substantial decrease in postpran- clinical trials have proven that intensive glycaemic dial hyperglycaemia in healthy individuals, as well as control can be achieved with insulin therapy in both in patients with type 1 or type 2 diabetes after sc or type 1 [1] and type 2 diabetes [2, 31]. intravenous administration [101]. Pramlintide was ap- The most commonly used human insulin prepara- proved in the US market in 2005 for the treatment of tions, all of which differ in their pharmacokinetics, type 1 and 2 diabetes in conjunction with insulin. The are summarized in Table 17.3. Hypoglycaemia remains Q half-life of pramlintide in healthy subjects is 48 min, the most frequently encountered side effect of insulin and the drug is metabolized and excreted primarily via therapy, and is the major limiting factor in intensive the kidneys. Patients with moderate or severe renal im- glycaemic control [1, 2]. Hypoglycaemia is commonly pairment (creatinine clearance >20 to <50 ml min−1) precipitated by erratic meal timing, excessive insulin did not show an increase in pramlintide exposure, or re- dosage and unplanned exercise. The failure to admin- duced clearance. The pramlintide starting dose is 15 μg ister regular insulin in a timely manner (30–45 min prior to the major meals, and this must be titrated every before a meal) also increases the risk. In addition to 3–7 days to a maximum of 60 μg before meals in type 1 the increased risk of hypoglycaemia, the recommended diabetes, and up to 120 μg in type 2 diabetes. The most timing of the injection of short-acting insulin prior to common adverse side effects include nausea, headache, meals can be inconvenient and may disrupt the pa- anorexia, vomiting and abdominal pain. Pramlintide tient’s lifestyle. alone does not cause hypoglycaemia, although coad- Recombinant DNA technology has been used to ministration of the drug with insulin can aggravate the design insulin molecules that overcome the limitations risk of insulin-induced hypoglycaemia, especially in of short-acting regular insulin. Yet, safety issues are a type 1 diabetes. With the initiation of pramlintide ther- concern with these alternatives, as the alteration of the apy, the insulin dose will need to be reduced in order three-dimensional structure may modify the interaction to prevent severe hypoglycaemic episodes. with the insulin and/or IGF-I receptors and, as a result, In a 52-week, double-blind, placebo-controlled study lead to the activation of alternative metabolic and of 656 patients with type 2 diabetes, treatment with mitogenic signalling pathways. It is, then, essential to pramlintide 120 μg b.i.d. led to a sustained reduction apply cautious study to the acute and long-term effects from baseline in HbA1c of −0.62%, and this was ac- in a preclinical state, as insulin therapy is meant to be companied by a mean body weight loss of −1.4 kg a lifelong treatment. compared to +0.7 kg with placebo [102]. In a simi- Ultra-short-acting insulin analogues were originally lar 1-year study of 651 type 1 diabetic patients, the developed with the hope of overcoming these limi- addition of pramlintide 60 μg t.i.d or q.i.d. to insulin tations. Lispro insulin is identical to human regular led to significant reductions in HbA1c from a base- insulin, but with a minor transposition of a lysine and line of 0.29% (P < 0.011) and 0.34% (P < 0.001), proline in the beta chain. This transposition results respectively, compared to a 0.04% reduction in the in an acceleration of the dissociation rate of the in- placebo group. This HbA1c reduction occurred with- sulin hexamers, such that lispro insulin acts quickly out any increase in concomitant insulin use, and was within 10–20 min, peaks on average at 1–2 h, and is accompanied by a significant reduction in body weight essentially cleared from the system within 4–5 h [104]. from baseline to week 52 of 0.4 kg in the 60 μg t.i.d. Thus, lispro provides a greater flexibility in insulin 256 CH 17 DRUG THERAPY: CURRENT AND EMERGING AGENTS FOR HYPERGLYCAEMIA

Table 17.3 Some of the most commonly used insulin preparations. Insulin preparations Action proﬁle (h) Onset Peak Duration Ultra-Rapid acting analogues (Human) Lispro 0.25–0.5 0.5–2 3–4 Aspart 0.25–0.5 0.5–2 3–4 Glulisine 0.25–0.5 0.5–2 3–4 Short-acting insulin (Human) Regular 0.5–1 2–3 6–8 U-500 1–3 6–12 12–18 Short-acting inhaled insulin Exubera 0.16–0.3 2 5–6 Intermediate-acting (Human) NPH 1.5 4–10 16–24 Intermediate-acting analogues Detemir 1–2 6–8 14–24 Long-acting analogues Glargine 1–1.5 No peak 20–24 Mixtures∗ Humulin 70/30 NPH 70%, regular 30% 0.5–1 3–12 16–24 Humulin 50/50 NPH 50%, regular 50% 0.5–1 2–12 16–24 Mixtures analogues∗ Humalog 75/25 NPL 75%, Lispro 25% 0.3–0.5 1–6.5 Up to 24 Humalog 50/50 NPL 50%, Lispro 50% 0.3–0.5 0.8–4.8 Up to 24 Novolog 70/30 NPL 70%, Aspart 30% 0.3–0.5 1.6–3.2 Up to 24

∗Mixtures with different proportions of NPH and regular, and more recently a mixture of lispro insulin and its protamine derivative, are also available but less commonly used.

administration, and can be given 5–10 min before a Many attempts have been made to identify an al- meal. Another modest advantage of lispro insulin has ternate method of insulin delivery, through nasal, oral, been the improvement in postprandial hyperglycaemia buccal, transdermal or inhaled routes. Inhaled insulin in both type 1 and type 2 diabetic subjects, and a reduc- is a theoretically attractive approach, mainly because tion in the number of hypoglycaemic episodes in type of the large absorptive surface of the lung alveoli 1 subjects [104, 105]. However, this latter advantage with relatively little proteolytic activity to cause in- over regular insulin has not been consistently demon- sulin degradation. One such product, Exubera (Pfizer), strated in type 2 subjects [104]. Overall, glycaemic has already been granted approval for clinical use control as reflected by HbA1C levels does not differ be- in the United States and the European Union, and tween lispro insulin and human regular insulin-treated other products [AERx (Novo Nordisk) and AIR (Eli groups [106]. Lilly)] are currently in the advanced stages of clini- Two other short-acting insulin analogues are avail- cal trials. In clinical studies in patients with type 1 able currently. ‘Insulin aspart’ (B28 Asp) has a proline and type 2 diabetes, following the inhalation of Exu- at position B28 replaced with a negatively charged bera, serum insulin reached a peak concentration more aspartic acid [107, 108]. Insulin glulisine has an as- quickly than after sc injection of regular human in- paragine replaced with lysine at position B2 and lysine sulin, namely 49 min (range 30–90 min) compared to with glutamic acid at B29 [108, 109]. In addition to 105 min (range 60–240 min). With inhaled insulin, the their rapid onset of action, insulin aspart and glulisine onset of glucose-lowering activity in healthy volunteers also have a favourable effect on postprandial hyper- occurred within 10–20 min, with the maximum effect glycaemia in both type 1 and type 2 diabetic subjects, on glucose lowering being exerted approximately 2 h while their pharmacokinetics resemble that of lispro after inhalation. The duration of the glucose-lowering insulin. activity was approximately 6 h. There were no ap- The clinical utility of these short-acting insulin ana- parent differences in the pharmacokinetic properties logues and their advantages over the regular insulin in of Exubera when comparing patients aged >65 years the elderly population has yet to be established. with younger, adult patients [110]. There is little doubt 17.10 THE CHOICE OF AN ANTI-DIABETIC AGENT IN ELDERLY PEOPLE 257 that treatment with inhaled insulin is associated with There is currently no strong rationale favouring a small (ca. 5%) decline in lung function, when mea- glargine, neutral protamine of Hagedorn insulin, insured by parameters such as the diffusion capacity of sulin detemir or fixed-ratio insulin preparations as the the lung, and forced expiratory volume in 1 s. The re- preferred agent for initiating insulin therapy. Rather, sults of a recent study indicated that treatment group an insulin regimen should be tailored to each patient’s differences in lung function between inhaled and sc in- needs [114]. sulin in adult patients with type 1 diabetes were small, developed early, and were non-progressive for up to 2 years of therapy [111]. There is a theoretical concern 17.10 The choice of an anti-diabetic for an increased risk of cancer with the long-term use agent in elderly people of inhaled insulin, since it is a potent growth factor. It might, therefore, be prudent to monitor for this po- In general, the management of type 2 diabetes should tential serious side effect that may take many years to be individualized and tailored to the clinical status emerge. of the individual, coexisting diseases, body weight, Longer-acting insulin preparations are prepared with goals of therapy, expectations, involvement in care, different chemical modifications of insulin. A pro- functional impairment, ease of administration, side tamine derivative of lispro insulin (NPL) is currently effect profile, cost of therapy, baseline blood sugar and available that has an intermediate duration of action. the urgency of blood sugar normalization. Although When NPL was premixed with lispro in a ratio of difficult to generalize when considering all these 75:25%, it was shown to improve postprandial gly- parameters, some guidelines can be suggested based caemic control compared to premixed human NPH and on blood sugar levels and the known clinical efficacy regular insulin in a 30:70% ratio [112]. and tolerability of the various classes of anti-diabetic Other intermediate- and long-acting insulin ana- agents. (Table 17.2) logues include glargine and detemir. Insulin glargine As in middle-aged individuals, lifestyle modifica- differs from human insulin in that the amino acid tion, including diet-control and exercise, remain the asparagine at position A21 is replaced by glycine, cornerstone of every treatment plan for type 2 dia- and two arginines are added to the C-terminus of the betes. However, under-nutrition is common in elderly B-chain. Insulin glargine is a human insulin analogue patients with type 2 diabetes, especially in those who that has been designed to have low aqueous solubility live in nursing homes [23, 116]. In this subgroup, the at neutral pH. After injection into the subcutaneous patient’s nutritional status should be carefully evaluate tissue, the acidic solution is neutralized, leading to and an appropriate caloric consumption recommended. the formation of microprecipitates from which small Elderly subjects also have more difficulties with the ac- amounts of insulin glargine are slowly released; this re- quisition, preparation and sometimes ingestion of food, sults in a relatively constant concentration–time profile and this can affect their nutritional status and increase over 24 h, with no pronounced peak. This pharmacoki- the risk of hypoglycaemic episodes [117]. Here, a sup- netic profile of glargine allows once-daily dosing as port system that can provide the appropriate meals in a the patient’s basal insulin [113]. Compared to bedtime timely manner is helpful. When there is no major phys- NPH insulin, insulin glargine is associated with less ical limitation, physical activity and exercise should be nocturnal hypoglycaemia in patients with type 2 dia- encouraged within reason. On the other hand, a care- betes (28.8% versus 12.6%, P = 0.011) [114]. ful medical evaluation – and sometimes also a limited Insulin detemir differs from human insulin in that stress test – should be undertaken before starting any the amino acid threonine in position B30 has been exercise programme, due to the high incidence of sub- omitted, and a C14 fatty acid chain has been attached clinical coronary artery disease in this population. to the amino acid B29. Insulin detemir is a soluble, Although it should be individualized, the trigger long-acting basal human insulin analogue with a rela- point to start pharmacological intervention in the el- tively flat action profile. The mean duration of action derly is not generally different from that in middle-aged of insulin detemir ranged from 5.7 h at the lowest dose diabetic people [118]. The goal for glycaemic control to 23.2 h at the highest dose, and the serum Cmax in the elderly diabetic is to achieve euglycaemia was reached at 6–8 h after administration [115]. Note: without the undue risk of hypoglycaemia. Ideally, Glargine and detemir should not be diluted or mixed the following blood glucose values should be sought: with any other insulin preparations. A FBG of 4.4–6.6 mmol l−1 (80–120 mg dl−1);a1-h 258 CH 17 DRUG THERAPY: CURRENT AND EMERGING AGENTS FOR HYPERGLYCAEMIA

postprandial glucose <8.8 mmol l−1 (160 mg dl−1); increased risk of atherosclerotic heart disease. and HbA1C <7%. Additional action is suggested if the Postprandial hyperglycaemia is probably one of FBG is >7.8 mmol l−1 (140 mg dl−1), the postprandial the reasons why some individuals with a FBG −1 −1 −1 glucose is >10 mmol l (180 mg dl ),ortheHbA1C within the target range (4.4–6.6 mmol l ) have is >7%. In order to avoid the deleterious effect of already established microvascular and macrovascular hypoglycaemia in the frail elderly diabetic patient, no complications. attempt should be made to lower the FBG or bedtime Despite the fact that non-pharmacological interven- blood glucose below 5.5 mmol l−1 (100 mg dl−1). tions alone may be sufﬁcient at early stages of the Although many physicians have traditionally relied disease, these measures will fail in most patients as only on the FBG and the HbA1C measurement to guide the disease progresses [2, 31]. An initial approach to pharmacological therapy, postprandial hyperglycaemia drug therapy of type 2 diabetes in elderly people is il- is becoming an important target of the management lustrated in Figure 17.2. When the FBG levels are con- [7, 36]. The rationale behind this approach is that sistently over 16.7 mmol l−1 (300 mg dl−1), it is most self-monitoring FBG and postprandial glucose are likely that the patient has either a profound insulin de- readily available for patients and clinicians, and can ﬁciency or severe insulin resistance, or both. Insulin be used to expedite the titration and changes in treatment – at least initially – is the preferred choice drug therapy. In addition, it appears that individuals for these individuals. Insulin therapy should be initi- with isolated postprandial hyperglycaemia are at ated in a hospital setting if the patient is symptomatic

Management of Type 2 Diabetes in the elderly

Diet & Exercise

Fasting Blood Glucose

>16.7 11.1-16.7 7.8-11.1 <7.8

Start basal Metformin* insulin: NPH , detemir Metformin*+ SFU Metformin* low dose SFU or glargine

Insulin Resistance PPG > 10 PPG < 10 PPG >10 PPG < 10 PPG >10 PPG <10

No Yes TZD Meglitinide Continue TZD Continue continue TZD/Metformin* Incretin- DPP-IV the same AGI the same the same Mimetic** AGI

PPG < 10 PPG >10 PPG < 10

Continue Add short acting Continue the same insulin analogues the same before meals or Stop basal insulin & start premixed BID

Figure 17.2 A suggested algorithm for the initial choice of therapeutic agents based on fasting blood glucose (in mmol l−1). The initial choice might be only for a brief period of time. PPG is 1 h postprandial blood glucose (in mmol l−1). SFU = sulphonylurea; AGI = alpha-glucosidase inhibitor; TZD = thiazolidinediones; DPP-IV = dipeptidyl-peptidase IV inhibitors; BID = twice daily. ∗If metformin is contraindicated, consider low-dose SFU, AGI, DPP-IV or TZD. ∗∗Adding insulin at this stage might be more cost-effective. 17.10 THE CHOICE OF AN ANTI-DIABETIC AGENT IN ELDERLY PEOPLE 259 with mental status changes or severe dehydration; oth- elderly, has any potential advantages over insulin ther- erwise, it can be started in the doctor’s surgery. The apy alone. When using such a regimen, it is important dose will be titrated during follow-up visits, and if there to consider the high cost of therapy and the possible is any evidence of significant insulin resistance then increased adverse events as a result of drug–drug in- insulin sensitizers, such as thiazolidinediones or met- teractions. This latter concern is very legitimate when formin, can be added if there are no contraindications. dealing with a frail population with multiple coexist- It is noteworthy, however, that some individuals (and ing medical conditions, and where polypharmacy is especially obese subjects) who present with a FBG common. Further studies are required to evaluate the >16.7 mmol l−1 may only require insulin for a short advantages and disadvantages of triple or quadruple period until the glucose toxicity resolves and glycaemic oral therapy in the elderly diabetic people and its im- goals can then be achieved by switching to oral agents. pact on the quality of life when compared to insulin When attending to with lean patients with an av- monotherapy. erage FBG of 11.1–16.7 mmol l−1 (200–300 mg dl−1), Until more data are available, the primary approach the initial drug of choice can be either a sulphonylurea is to tailor the management to individual needs. Dif- or insulin. For obese subjects (i.e. BMI ≥30 kg m−2), ferent individuals require different drug regimens, and metformin would be added, although due to a high this may change over time as the disease progresses. prevalence of contraindications to biguanides in the el- It should be borne in mind that type 2 diabetes is a derly, TZD – particularly pioglitazone – might be an heterogeneous disease. Now that multiple pharmaco- alternative to metformin. It is important that the po- logical agents are available, with distinct mechanisms tential for cardiovascular events associated with these of action, it should be more possible than ever to in- agents is thoroughly discussed with the patients. If the dividualize management by matching the appropriate FBG is within target, but postprandial hyperglycaemia agent to the underlying pathophysiology of type 2 di- remains a problem, then consideration should be given abetes. to adding a GLP-1 mimetic or initiating insulin therapy. For individuals with a FBG of 7.8–11.1 mmol l−1 (140–200 mg dl−1), metformin should be the initial References drug and subsequently a low dose of sulphonylurea can be added. Alternatively, a TZD might be added or 1. The Diabetes Control and Complications Trial Re- search Group (1993) The effect of intensive treatment started, if metformin were contraindicated. of diabetes on the development and progression of The cost of a triple oral therapy is of concern (as are long-term complications in insulin-dependent diabetes drug–drug interactions), and this should probably be mellitus. N Engl J Med, 329, 977–86. reserved for those individuals who refuse insulin ther- 2. UK Prospective Diabetes Study Group 33 (1998) −1 −1 apy. When the FBG is <7.8 mmol l (140 mg dl ),a Intensive blood-glucose control with sulfonylureas or trial of an exercise programme with a weight-reducing insulin compared with conventional treatment and diet for a 2-month period is initially appropriate. Sub- risk of complications in patients with type 2 diabetes sequently, if the glycaemic goals are not within target, (UKPDS 33). Lancet, 352, 837–53. then metformin should be the initial drug and an AGI, 3. Morley JE, Mooradian AD, Rosenthal MJ and Kaiser DPP-IV inhibitor or meglitinide added as appropriate. FE (1987) Diabetes mellitus in elderly patients: is it The addition of a sulphonylurea is definitely more cost- different? Am J Med, 83, 533–44. effective, but the risk of hypoglycaemia will be in- 4. Rosenthal MJ and Morley JE (1992) Diabetes and creased. Alternatively, consideration could be given to its complications in older people. In: Endocrinology adding or starting with a thiazolidinedione, especially and Metabolism in the Elderly, JE Morley and SG when metformin is contraindicated Korenman (eds), Blackwell Scientific Publications, Boston, pp. 373–87. One common problem in clinical practice is that 5. Oiknine R and Mooradian AD. (2003) Drug therapy monotherapy with oral agents either fails initially, or of diabetes in the elderly. Biomed Pharmacother. 57, fails over time. This important clinical observation was 231–9. confirmed in the UKPDS. In order to achieve an opti- 6. Mooradian AD. (2004) Towards single-tablet therapy mal outcome, multiple drug therapy or a combination for type 2 diabetes mellitus: rationale and recent of oral agents and insulin is often needed. An important developments. Treat Endocrinol. 3, 279–87. question that must be explored here is whether triple 7. Mooradian AD and Thurman J (1999) Drug therapy oral therapy, when used to avoid insulin use in the of postprandial hyperglycemia. Drugs, 97 (1), 19–29. 260 CH 17 DRUG THERAPY: CURRENT AND EMERGING AGENTS FOR HYPERGLYCAEMIA

8. Groop LC (1992) Sulfonylureas in NIDDM. Diabetes 22. Rosenkranz B, Profozic V, Metelko Z, Mrzljak V, Care, 19, 737–54. Lange C and Malerczyk V (1996) Pharmacokinetics 9. Pfeiffer MA, Beard JC, Halter JB, Judzewitsch R, and safety of glimepiride at clinically effective doses Best JD and Porte D Jr. (1983) Suppression of in diabetic patients with renal impairment. Diabetolo- glucagon secretion during a tolbutamide infusion in gia, 39, 1617–24. normal and noninsulin-dependent diabetic subjects. J 23. Mooradian AD, Osterweil D, Petrasek D and Morley Clin Endocrinol Metab, 56, 586–91. JE (1988) Diabetes mellitus in elderly nursing home 10. Keller U, Muller R and Berger W (1986) Sulfonylurea patients: a survey of clinical characteristics and man- therapy fails to diminish insulin resistance in type 1 agement. J Am Geriatr Soc, 36, 391–6. diabetic subjects. Horm Metab Res, 18, 599–603. 24. Horani MH and Mooradian AD. (2002) Management 11. Kolterman OG and Olefsky JM (1984) The impact of obesity in the elderly. Special considerations. Treat- of sulfonylurea treatment upon the mechanisms re- ments in Endocrinology, 1, 387–98. sponsible for the insulin resistance in type II diabetes. 25. Bailey CJ (1992) Biguanides and NIDDM. Diabetes Diabetes Care, 7 (Suppl. 1), 81–8. Care, 15, 755–72. 12. Mooradian AD (1987) The effect of sulfonylureas on 26. Williams RH and Palmer JP (1975) Farewell to phen- the in vivo tissue uptake of glucose in normal rats. formin for treating diabetes mellitus. Ann Intern Med, 83, 567–8. Diabetologia, 30, 120–1. ® 13. Groop L, Groop PH, Stenman S, Saloranta C, 27. Bristol-Myers Squibb (1995) Glucophage (met- Totterman¨ KJ, Fyhrquist F and Melander A (1988) formin hydrochloride tablets) 500 mg and 850 mg. Do sulfonylureas influence hepatic insulin clearance? Prescribing information. Princeton, New Jersey: Diabetes Care, 11, 689–90. Bristol-Myers Squibb Company. 14. Beck-Nielsen H, Hother Nielsen O, Andersen PH, 28. Bailey CJ and Turner RC (1996) Metformin. New Engl J Med, 334, 574–9. Pederson O and Schmitz O (1986) In vivo action 29. Perriello G, Misericordia P, Volpi E, Santucci A, San- of glibenclamide. Diabetologia, 29 (Abstract No. 26), tucci C, Ferrannini E, Ventura MM, Santeusanio F, 515A. Brunetti P and Bolli GB (1994) Acute antihyper- 15. Kahn CR and Schecte Y (1993) Oral hypoglycemic glycemic mechanisms of metformin in NIDDM: evi- agents. In: The Pharmacologic Basis of Therapeutics, dence for suppression of lipid oxidation and hepatic AG Gilman, TW Rall, AS Nies, P Taylor (eds), glucose production. Diabetes, 43, 920–8. McGraw-Hill, New York, p. 1485. 30. Garber AJ, Duncan TG, Goodman AM, Mills DJ and 16. Melander A, Bitzen´ PO, Faber O and Groop L (1989) Rohlf JL (1997) Efficacy of metformin in type II di- Sulphonylurea antidiabetic drugs: an update of their abetes: Results of a double-blind, placebo-controlled, clinical pharmacology and rational therapeutic use. dose response trial. Am J Med, 102, 491–7. Drugs, 37, 58–72. 31. UK Prospective Diabetes Study Group 34 (1998) 17. Scheen AJ (1997) Drug treatment on Effect of intensive blood-glucose control with met- non-insulin-dependent diabetes mellitus in the formin on complications in overweight patients with 1990s. Achievements and future developments. Type 2 diabetes (UKPDS 34). Lancet, 352, 854–65. Drugs, 54 (3), 355–68. 32. Yki-Jarvinen H, Ryysy L, Nikkila¨ K, Tulokas T, 18. Groop L, Schalin C, Franssila-Kallunki A, Widen´ Vanamo R and Heikkila M (1999) Comparison of E, Ekstrand A, Eriksson J. Characteristics of bedtime insulin regimens in patients with Type 2 non-insulin-dependent diabetic patients with diabetes mellitus. A randomized controlled trial. Ann secondary failure to oral antidiabetic therapy. Am J Intern Med, 130, 389–96. Med. 1989; 87 (2): 183–90. 33. Defronzo RA and Goodman AM, (1995) Multicen- 19. Jennings AM, Wilson RM and Ward JD (1989) Symp- ter Metformin Study Group. (1995) Efficacy of met- tomatic hypoglycemia in NIDDM patients treated formin in patients with non-insulin diabetes mellitus. with oral hypoglycemic agents. Diabetes Care, 12, N Engl J Med, 333, 541–9. 203–8. 34. Stumvoll M, Nurjhan N, Perriello G, Dailey G and 20. UK Prospective Diabetes Study Group 16 (1995) Gerich JE (1995) Metabolic effects of metformin U.K. Prospective Diabetes Study. Overview of 6years’ in non-insulin-dependent diabetes mellitus. N Eng J therapy of Type II diabetes: a progressive disease. Med, 333, 550–4. Diabetes, 44, 1249–58. 35. Grosskopf I, Ringel Y, Charach G, Maharshak 21. Asplund K, Wiholm BE and Lithner F (1983) N, Mor R, Iaina A and Weintraub M (1997) Glibenclamide-associated hypoglycemia: a report of Metformin enhances clearance of chylomicrons 57 cases. Diabetologia, 24, 412–17. and chylomicron remnants in nondiabetic mildly 17.10 THE CHOICE OF AN ANTI-DIABETIC AGENT IN ELDERLY PEOPLE 261

overweight glucose-intolerant subjects. Diabetes 49. Baron A and Neumann C (1997) PROTECT interim Care, 20, 1598–602. results: A large multicenter study of patients with type 36. Mooradian AD (1996) Drug therapy of II diabetes. Clin Ther, 19, 282–95. non-insulin-dependent diabetes mellitus in the 50. Baron A and Neumann C. PROTECT interim results: elderly. Drugs, 6, 931–41. a large multicenter study of patients with type II 37. Bergman U, Boman G and Wilholm BE (1978) Epi- diabetes. Precose Resolution of Optimal Titration to demiology of adverse drug reactions to phenformin Enhance Current Therapies. Clin Ther. 1997; 19 (2): and metformin. Br Med J, 2 (6135), 464–6. 282–95. 38. Stang M, Wysowski DK and Butler- Jones D (1999) 51. Chiasson JL, Josse RJ, Gomis R, Hanefeld M, Karasik Incidence of lactic acidosis in metformin users. Dia- A, Laakso M and the STOP-NIDDM TRIAL research betes Care, 22, 925–7. group (2003) Acarbose treatment and the risk of 39. Rabasa-Lhoret R and Chiasson JL (1998) Potential of cardiovascular disease and hypertension in patients α-glucosidase inhibitors in elderly patients with dia- with impaired glucose tolerance: the STOP-NIDDM betes mellitus and impaired glucose tolerance. Drugs trial. JAMA, 290(4), 486–94. and Aging, 13 (2), 131–43. 52. Johnston PS, Lebovitz HE, Coniff RF, Simonson DC, 40. Clissold SP and Edwards C (1988) Acarbose. A pre- Raskin P and Munera CL (1998) Advantages of alpha liminary review of its pharmacodynamic and pharma- glucosidase inhibition as monotherapy in elderly type cokinetic properties, and therapeutic potential. Drugs, 2 diabetic patients. J Clin Endocrinol Metab, 83, 35, 214–43. 1515–22. 41. Santeusanio F and Compagnucci P (1994) A 53. Mooradian AD, Thurman JE and Chehade JM (2002) risk-benefit appraisal of acarbose in the management The role of thiazolidinediones in the treatment of pa- of non-insulin-dependent diabetes mellitus. Drug tients with type 2 diabetes mellitus. Treat Endocrinol. Safety, 11, 432–44. 1, 13–20. 42. Bayer Corporation (1998) Precose® (acarbose). Pre- 54. SmithKline Beecham (1999) Avandia (Rosiglitazone) scribing information. Bayer Corporation, West Haven, tablets. Prescribing information. Philadelphia, Penn- CT, USA. sylvania, USA. 43. Deleon MJ, Chandurkar V, Albert SG and Moora- 55. Maggs DG, Buchanan TA, Burant CF, Cline G, dian AD. (2002) Glucagon-like peptide-1 (GLP-1) re- Gumbiner B, Hsueh WA, Inzucchi S, Kelley D, Nolan sponse to acarbose in elderly type 2 diabetic subjects. J, Olefsky JM, Polonsky KS, Silver D, Valiquett TR Diabetes Research and Clinical Practice, 56, 101–6. and Shulman GI (1998) Metabolic effects of trogli- 44. Pharmacia and Upjohn 1998. Glyset® (miglitol). Pre- tazone monotherapy in Type 2 diabetes mellitus. A scribing information. Pharmacia & Upjohn Company, Kalamazoo, MI, USA. randomized double blind placebo controlled trial. Ann 45. Chiasson JL, Josse RG, Hunt JA, Palmason C, Rodger Intern Med, 128, 176–85. NW, Ross SA, Ryan EA, Tan MH and Wolever TM 56. Godlberg RB, Kendall DM, Deeg MA, Buse JB, (1994) The efficacy of acarbose in the treatment of Zagar AJ, Pinaire JA, Tan MH, Khan MA, Perez patients with non-insulin dependent diabetes mellitus. AT, Jacober SJ and the GLAI Study investigators A multicenter controlled clinical trial. Ann Intern (2005). A comparison of lipid and glycemic effects Med, 121, 928–35. of pioglitazone and rosiglitazone in patients with type 46. Holman RR, Cull CA and Turner RC (1999) A ran- 2 diabetes and dyslipidemia. Diabetes Care, 28 (7), domized double-blind trial of acarbose in type 2 dia- 1547–54. betes shows improved glycemic control over 3years. 57. Schnieder R, Egan J and Houser V (1999) Piogli- (U.K. Prospective Diabetes Study 44). Diabetes Care, tazone 101 Study Group. Combination therapy with 22, 960–4. pioglitazone and sulfonylurea in patients with type 2 47. Mooradian AD, Albert SG, Wittry S, Chehade J, Kim diabetes. Diabetes, 48 (Suppl. 1), Abstract 458 A106. J and Bellrichard B. (2000) Dose-response profile of 58. Gomis R, Jones NP, Vallance SE and Ratwardhan acarbose in older subjects with Type 2 diabetes. Am. R (1999) Low dose rosiglitazone provides additional J. Med. Sci., 319, 334–7. glycemic control when combined with sulfonylureas 48. Johnston PS, Lebovitz HE and Coniff R (1997) Ad- in type 2 diabetes. Diabetes, 48 (Suppl. 1), Poster 266 vantages of monotherapy with alpha-glucosidase in- A63. hibitors in elderly NIDDM patients (abstract). Dia- 59. Promrat K, Lutchman G, Uwaifo GI, Freedman RJ, betes, 46 (Suppl. 1), 158A. Soza A, Heller T, Doo E, Ghany M, Premkumar A, 262 CH 17 DRUG THERAPY: CURRENT AND EMERGING AGENTS FOR HYPERGLYCAEMIA

Park Y, Liang TJ, Yanovski JA, Kleiner DE and Hoof- 72. Fonseca V, Grunberger G, Gupta S, et al. (2003) nagle JH (2004) A pilot study of pioglitazone treat- Addition of Nateglinide to Rosiglitazone monother- ment for nonalcoholic steatohepatitis. Hepatology, 39 apy suppresses mealtime hyperglycemia and improves (1), 188–96. overall glycemic control. Diabetes Care, 26, 1685–90. 60. Neuschwander-Tetri BA, Brunt EM, Wehmeier KR, 73. Jovanovic L, Hassman DR, Gooch B, et al. (2004) Oliver D and Bacon BR (2003) Improved nonalco- Treatment of type 2 diabetes with a combination holic steatohepatitis after 48 weeks of treatment with regimen of repaglinide plus pioglitazone. Diab Res the PPAR-gamma ligand rosiglitazone. Hepatology, Clin Pract, 63, 127–34. 74. Tornier B, Marbury TC, Dambso P and Windﬁeld K 38 (4), 1008–17. (1995) A new oral hypoglycemic agent, Repaglinide, 61. Nissen SE and Wolski K (2007) Effect of rosiglitazone minimizes risk of hypoglycemia in well controlled on the risk of myocardial infarction and death from Type 2 diabetic patients. Diabetes, 44 (Suppl. 1), 70A, cardiovascular causes. N Engl J Med, 356, 2457–71. Abstract. 62. Dormandy JA et al. and the PROactive investiga- 75. Mooradian AD (1998) Repaglinide: A viewpoint. tors. Secondary prevention of macrovascular events Drugs and Aging, 13 (2), 181. in patients with type 2 diabetes in the PROactive 76. Elrick H, Stimmler L, Hlad CJ and Arai Y (1964) Study (PROspective pioglitAzone Clinical Trial In Plasma insulin response to oral and intravenous glu- macroVascular Events): a randomised controlled trial. cose administration. J Clin Endocrinol Metab, 24, Lancet, 2005; 366 (9493): 1279–89. 1076–82. 63. Malaisse WJ. Stimulation of insulin release 77. Holst JJ (1994) Glucagon-like peptide-1 (GLP-1) - a by non-sulfonylurea hypoglycemic agents: the newly discovered gastrointestinal hormone. Gastroen- meglitinide family. Horm Metab Res. 1995; 27 (6): terology, 107, 1848–55. 263–6. 78. Ørskov C (1992) Glucagon-like peptide-1, a new 64. Balfour JA and Faulds D (1998) Repaglinide. Drugs hormone of the enteroinsular axis. Diabetologia, 35, and Aging, 13, 173–80. 701–11. 79. Drucker DJ, Philippe J, Mozsov S, Chick W and 65. Novo Nordisk (1997) Prandin® (repaglinide). Pre- Habener JF (1987) Glucagon-like peptide I stimulates scribing information. Novo Nordisk Pharmaceuticals, insulin gene expression and increases cyclic AMP Princeton, NJ, U.S.A. levels in a rat islet cell line. Proc Natl Acad Sci USA, 66. McLeod JF. Clinical pharmacokinetics of Nateglinide 84, 3434–8. (2004) A rapidly-absorbed, short-acting insulinotropic 80. Goke¨ R, Wagner B, Fehmann HC and Goke¨ B (1993) agent. Clin Pharmacokinet, 43, 97–120. Glucose-dependency of the insulin stimulatory effect 67. Hatorp V (2002) Clinical pharmacokinetics and phar- of glucagon-like peptide-1 (7-36) amide on the rat macodynamics of repaglinide. Clin Pharmacokinet., pancreas. Res Exp Med, 193, 97–103. 41 (7), 471–83. 81. Willms B, Werner J, Holst JJ, Orskov C, Creutzfeldt 68. Marbury TC, Ruckle JL, Hatorp V, Andersen MP, W and Nauck MA. Gastric emptying, glucose Nielsen KK, Huang WC and Strange P (2000) Phar- responses, and insulin secretion after a liquid macokinetics of repaglinide in subjects with renal test meal: effects of exogenous glucagon-like impairment. Clin Pharmacol Ther., 67(1), 7–15. peptide-1 (GLP-1)-(7-36) amide in type 2 69. Derosa G, Mugellini A, Ciccarelli L, et al. (2003) (noninsulin-dependent) diabetic patients. J Clin Comparison between repaglinide and Glimepiride in Endocrinol Metab. 1996; 81 (1): 327–32. patients with type 2 diabetes mellitus: a one-year, 82. Gutniak MK, Larsson H, Sanders SW, Juneskans randomized, double-blind assessment of metabolic pa- O, Holst JJ and Ahren´ B (1997) GLP-1 tablet in type 2 diabetes in fasting and postprandial conditions. rameters and cardiovascular risk factors. Clin Thera- Diabetes Care, 20 (12), 1874–9. peutics, 25, 472–84. 83. Parkes DG, Pittner R, Jodka C, Smith P and 70. Marbury T, Huang WC, Strange P, Lebovitz H (1999) Young A (2001) Insulinotropic actions of exendin-4 Repaglinide versus glyburide: a one-year comparison and glucagon-like peptide-1 in vivo and in vitro. trial. Diabetes Res Clin Pract., 43 (3), 155–66. Metabolism, 50 (5), 583–9. 71. Moses R, Slobodniuk R, Boyages S, Colagiuri S, 84. Amylin Pharmaceuticals 2005. Byetta (package in- Kidson W, Carter J, Donnelly T, Mofﬁtt P and Hopkin sert). Amylin, San Diego, CA. H (1999) Effect of repaglinide addition to metformin 85. DeFronzo RA, Ratner RE, Han J, Kim DD, Fine- monotherapy on glycemic control in patients with type man MS and Baron AD (2005) Effects of exenatide 2 diabetes. Diabetes Care, 22 (1), 119–24. (exendin-4) on glycemic control and weight over 30 17.10 THE CHOICE OF AN ANTI-DIABETIC AGENT IN ELDERLY PEOPLE 263

weeks in metformin-treated patients with type 2 dia- impairment of insulin secretion in Type II diabetes. betes. Diabetes Care, 28 (5), 1092–100. Diabetes, 40, 1615–19. 86. Buse JB, Henry RR, Han J, Kim DD, Fineman 97. Fineman MS, Giotta MP, Thompson RG, Kolterman MS, Baron AD and the Exenatide 113 Clini- OG and Koda JE (1996) Amylin response following cal Study Group (2004) Effects of exenatide Sustacal ingestion is diminished in type II diabetic (exendin-4) on glycemic control over 30 weeks in patients treated with insulin. Diabetologia, 39 (Suppl. sulfonylurea-treated patients with type 2 diabetes. 1), A149, Abstract. Diabetes Care, 27 (11), 2628–35. 98. Kolterman OG, Gottlieb A, Moyses C, Colburn W. 87. Kendall DM, Riddle MC, Rosenstock J, Zhuang D, Reduction of postprandial hyperglycemia in subjects Kim DD, Fineman MS and Baron AD (2005) Effects with IDDM by intravenous infusion of AC137, a of exenatide (exendin-4) on glycemic control over 30 human amylin analogue. Diabetes Care 1995; 18 (8): weeks in patients with type 2 diabetes treated with 1179–82. metformin and a sulfonylurea. Diabetes Care, 28 (5), 99. Gedulin BR, Rink TJ and Young AA (1997) 1083–91. Dose-response for glucagonostatic effect of amylin 88. Klonoff DC, Buse JB, Nielsen LL, Guan X, Bowlus in rats. Metabolism, 46, 67–70. CL, Holcombe JH, Wintle ME and Maggs DG. Exe- 100. Thompson RG, Gottlieb A, Organ K, Koda J, Kisicki natide effects on diabetes, obesity, cardiovascular risk J, Kolterman OG. Pramlintide: a human amylin ana- factors and hepatic biomarkers in patients with type logue reduced postprandial plasma glucose, insulin, 2 diabetes treated for at least 3years. Curr Med Res and C-peptide concentrations in patients with type 2 Opin. 2008; 24 (1): 275–86. diabetes. Diabet Med. 1997; 14 (7): 547–55. 89. Mentlein R (1999) Dipeptidyl-peptidase IV (CD-26)- 101. Thompson RG, Pearson L, Schoenfeld S and Kolter- role in the inactivation of regulatory peptides. Regul man OG (1998) The Pramlintide in Type 2 Dia- betes Group. Pramlintide, a synthetic analog of human Pept, 85, 9–24. amylin, improves the metabolic profile of patients 90. Drucker DJ and Nauck MA (2006) The incretin sys- with type 2 diabetes using insulin. Diabetes Care, 21 tem: glucagon-like peptide-1 receptor agonists and (6), 987–93. dipeptidyl peptidase-4 inhibitors in type 2 diabetes. 102. Hollander PA, Levy P, Fineman MS, Maggs DG, Shen Lancet, 368 (9548), 1696–705. LZ, Strobel SA, Weyer C and Kolterman OG (2003) 91. Nauck MA, Meininger G, Sheng D, Teranella L, Pramlintide as an adjunct to insulin therapy improves Stein P for the Sitagliptin Study Group (2007) Ef- long-term glycemic and weight control in patients ficacy and safety of the dipeptidyl peptidase-4 in- with type 2 diabetes: a 1-year randomized controlled hibitor, sitagliptin, compared with the sulfonylurea, trial. Diabetes Care, 26 (3), 784–90. glipizide, in patients with type 2 diabetes inade- 103. Ratner RE, Dickey R, Fineman MS, Maggs DG, Shen quately controlled on metformin alone: a random- LZ, Strobel SA, Weyer C and Kolterman OG (2004) ized, double-blind, non-inferiority trial. Diabetes Obes Amylin replacement with pramlintide as an adjunct Metab. 2, 194–205. to insulin therapy improves long-term glycaemic and 92. Amori RE, Lau J and Pittas AG (2007) Efficacy weight control in Type 1 diabetes mellitus: a 1-year, and safety of incretin therapy in type 2 diabetes: randomized controlled trial. Diabet Med., 21 (11), systematic review and meta-analysis. Lancet, 298 (2), 1204–12. 194–206. 104. Holleman F and Hoekstra JB (1997) Insulin Lispro. 93. Cooper GJS, Willis AC, Clark A, Turner RC, Sim N. Engl. J. Med., 337, 176–83. RB and Reid KB (1987) Purification and character- 105. Anderson JH, Jr, Brunelle RL, Keohane P, Koivisto ization of a peptide from amyloid-rich pancreas of VA, Trautmann ME, Vignati L and DiMarchi R type 2 diabetic patients. Proc Natl Acad Sci USA, 84, (1997) Mealtime treatment with insulin analog im- 8628–32. proves postprandial hyperglycemia and hypoglycemia 94. Inoue K, Hisatomi A, Umeda F and Nawata H (1991) in patients with non-insulin-dependent diabetes mel- Release of amylin from perfused rat pancreas in litus. Arch Intern Med, 157, 1249–55. response to glucose, arginine, beta-hydroxybutyrate, 106. Anderson JH, Jr, Brunelle RL, Koivisto VA, Pfutzner and gliclazide. Diabetes, 40, 1005–9. A, Trautmann ME, Vignati L and Dimarchi R 95. Koda JE, Fineman M, Rink TJ, Dailey GE, Muchmore (1997) Reduction of postprandial hyperglycemia and DB and Linarelli LG (1992) Amylin concentrations frequency of hypoglycemia in IDDM patients on and glucose control. Lancet, 339 (8802), 1179–80. insulin-analog treatment. Diabetes, 46, 265–70. 96. Ludvik B, Lell B, Hartter E, Schnack C and Prager R 107. Simpson KL and Spencer CM (1999) Insulin Aspart. (1991) Decrease of stimulated amylin release precedes Drugs, 57 (5), 759–65. 264 CH 17 DRUG THERAPY: CURRENT AND EMERGING AGENTS FOR HYPERGLYCAEMIA

108. Oiknine R, Bernbaum M and Mooradian AD. (2005) 113. Sanofi-Aventis (2000) Lantus® (insulin glargine A critical appraisal of the role of insulin analogues in [rDNA origin] injection). Prescribing information. the management of diabetes. Drugs, 65 (3), 325–40. Sanofi-Aventis, Bridgewater, NJ, USA 109. Becker RH (2007) Insulin glulisine complementing 114. Mooradian AD, Bernbaum M and Albert SG (2006) basal insulins: a review of structure and activity. A narrative review: a rational approach to starting Diabetes Technol Ther, 9 (1), 109–21. insulin therapy. Annal Intern Med, 145 (2), 125–34. ® 110. Pfizer (2006) EXUBERA (insulin human [rDNA 115. Novo Nordisk (2005) Levemir (insulin detemir origin]). Inhalation Powder US Package Insert. [rDNA origin] injection). Package insert, Novo 111. Skyler JS, Jovanovic L, Sol Klioze S, Reis J, Duggan Nordisk, Princeton, NJ. W, for the Inhaled Human Insulin Type 1 Diabetes 116. Mooradian AD, Kalis J and Nugent CA (1990) The Study Group (2007) Two-year safety and efficacy of nutritional status of ambulatory elderly type II diabetic inhaled human insulin (Exubera) in adult patients with patients. Age, 13, 87–9. type 1 diabetes. Diabetes Care, 30, 579–85. 117. Reed RL and Mooradian AD (1990) Nutritional status 112. Roach P, Yue L and Arora V (1999) Improved post- and dietary management of elderly diabetic patients. prandial glycemic control during treatment with Hu- Clin Geriatr Med, 6, 883–901. malog M; a novel protamine-based insulin lispro for- 118. Chehade JM and Mooradian AD (2000) A rational mulation. Humalog Mix25 Study. Diabetes Care, 22, approach to drug therapy of type 2 diabetes mellitus. 1258–61. Drugs, 60, 95–113. 18 Insulin Therapy

Dennis Yue and Marg McGill Diabetes Centre, Royal Prince Alfred Hospital, Camperdown, NSW, Australia

the majority of patients encountered. Such patients will Key messages usually provide a history of having diabetes for more than a few years and then gradually losing their re- • Insulin therapy may be required at some stage sponsiveness to oral hypoglycaemic agents, which is in up to 30% of older people, and is usually now known to be a consequence of the inevitable de- prescribed to optimize metabolic control. cline in pancreatic beta-cell function. This profile is so • Insulin regimens used in younger patients are characteristic that, from a clinical perspective, if an el- equally applicable to the elderly, although in derly patient has a short duration of diabetes and then rather frail or highly dependent patients, the rapidly becomes insulin requiring, one should consider regimen is usually once-daily insulin plus or minus tablet therapy. the presence of an underlying pancreatic carcinoma • Starting insulin therapy in older people requires causing insulin deficiency. the consideration of several patient-related fac- There is a general consensus worldwide that insulin tors, the likelihood of carer involvement, and therapy is not used early or often enough in the treat- the agreed goals of treatment. ment of diabetes. Many young patients languish for years on supra-maximal dosages of oral hypoglycaemic agents whilst hyperglycaemia insidiously but unrelent- ingly causes tissue damage and complications. The 18.1 Introduction failure to introduce insulin therapy sufficiently early is, in many countries, due to lack of resources, but in Diabetes is a condition that becomes more prevalent others it is due to inappropriate and entrenched atti- with age and, indeed, is one of the most common tudes of health professionals and people with diabetes, chronic diseases of elderly people [1]. It is inevitable alike. Obviously, such factors affecting the use of in- therefore that, as the population ages, the number of sulin in a particular country would also impact on the older people with diabetes will increase. Most people use of insulin therapy in the elderly in the community. with diabetes have the type 2 condition, and almost half However, some other factors may also play a role, in of all persons known to have type 2 diabetes are aged that the use of insulin in the elderly can vary consid- >65 years [2]. Occasionally, type 1 diabetes can appear erably in many developed countries of similar health for the first time in the elderly and, with better gen- care standards [4]. For example, in the United States eral medical care, many patients with type 1 diabetes insulin therapy is used in up to 25% of elderly people survive more than 50–60 years with the condition [3]. with diabetes, whereas in a population-based cohort of However, these scenarios are rare, and in the context persons aged ≥65 years in France, insulin was found to of starting insulin therapy, type 2 diabetes comprises be used by only 6.5% of such patients [5]. In contrast,

Diabetes in Old Age Third Edition Edited by Alan J. Sinclair © 2009 John Wiley & Sons Ltd. ISBN: 978-0-470-06562-4 266 CH 18 INSULIN THERAPY in United Kingdom nursing homes insulin was found US Veterans Health Administration guidelines suggest to be used in 25–47% of elderly people with type 2 aiming for a HbA1c of: diabetes [6]. ≥ In younger diabetic individuals whose glycaemic <7% if the person’s predicted lifespan is 15 years, control is not adequately achieved after an appropri- and there are no microvascular complications; or ≥ ate period of treatment with dietary measures and oral 10 years if early to moderate microvascular com- medications, the introduction of insulin therapy is al- plications are present. most always indicated and correct. This has the proven <8% if the life span is 5–15 years if no microvascular benefit of reducing the development and severity of complications are present, or 5–10 years if early to chronic diabetic complications, this being a matter of moderate microvascular complications are present. great importance not only to the individuals but also <9% if the lifespan is thought to be <5 years, with or to the nation from a health economics point of view. without microvascular complications. In the elderly, however, the same is not always true. Implicit in these guidelines is the fact that glycaemic Elderly patients can be unnecessarily commenced on goals should be individualized to the specific circum- insulin therapy, causing inconvenience to the person stances of the elderly, balancing the need to prevent and his/her family, with little possibility of improv- microvascular complications against life expectancy, ing health outcomes. Furthermore, as tight glycaemic largely governed by comorbidities such as stroke, can- control is often associated with an increased risk of cer and severe heart failure [9], all of which are less severe hypoglycaemia in the elderly, this can have se- amenable to the benefits of tight glycaemic control. rious consequences such as precipitating a fall with In this regard, the duration of diabetes should also be fractures, or exacerbating ischaemic heart disease. It is taken into consideration in making this decision. It is noteworthy that the median age of patients with severe true that many patients have hyperglycaemia well be- hypoglycaemia in Sweden was 75 years, and 21% of fore being discovered to have diabetes, and thus may these were aged >85 years, with an overall fatality rate have microvascular complications soon after presenta- of 20% [7]. On the other hand, elderly patients, even tion. However, by and large the development of mi- more than their younger counterparts, are often denied crovascular complications is still a function of known insulin therapy because of their age alone [6]. Thus, duration of diabetes. Therefore, the risk of developing when – and how – to start elderly patients on insulin severe microvascular complications in a patient diag- therapy is an important issue for which there is rela- nosed in their 70s or 80s is not high, provided that tively little evidence-based data to guide the clinicians. they do not already have evidence of complications at Rather, they must instead rely on their commonsense presentation. These individuals can afford to have gly- and clinical skills, not only as a health care provider caemic targets that are less tight compared to a person but often also as a psychologist and a social worker. who has been diagnosed for many years and is now in In the following section, we provide a perspective on their 70s, or older. what are becoming increasingly common issues in the practical care of the elderly person with diabetes. 18.3 Other indications for insulin 18.2 What glycaemic threshold therapy in the elderly should trigger insulin therapy It is important to recognize that the application of the in the elderly? above guidelines should be completely over-ridden by any need to use insulin therapy to control hypergly- The American Geriatric Society has published guide- caemic symptoms and infection. If an elderly person is lines regarding glucose control for older ambulatory experiencing distressing thirst, nocturia and either skin, adults with diabetes [8]. These guidelines acknowl- vaginal or penile monilial infection, then insulin ther- edged the possible risks associated with tight gly- apy is indicated irrespective of the patient’s age or life caemic control, and therefore suggested a modified expectancy [10]. Elderly patients often do not volun- HbA1c goal of <8%. Conceptually, this needs to teer these symptoms. Questions such as “Do you have be modified according to the status of microvascular to get up at night to go to the toilet?”, “Is it worse complications and life expectancy. On this basis, the than last year?” and “Do you have any itch in your 18.4 SOME OF SPECIAL CONSIDERATIONS OF ELDERLY PEOPLE AND INSULIN THERAPY 267 front passage?” are simple tools which cannot be read- A variety of data, albeit limited, has demonstrated ily replaced by investigations, and the response would that those being cared for in nursing homes often do greatly influence our decision to implement insulin not receive care of an adequate standard. As a conse- therapy. Other clinical parameters that can be added quence, many people unnecessarily live with elevated to this list include tiredness, poorer general health and blood glucose levels that interfere with their quality weight loss. Whilst the first two are harder to quantify, of life and make them more susceptible to infection a well-documented weight loss of several kilograms is and coma. One of the reasons for this may be the lack an important sign indicative of insulin deficiency and of nursing staff in some institutions, which makes the the need to commence insulin therapy, provided that delivery of insulin injections problematic when the ma- other causes seem unlikely. jority of staff are care assistants. A further contributor In assessing elderly people’s symptoms and to the situation is the staff’s fear of hypoglycaemia well-being with regards to insulin therapy, it is worth due to the unreliable eating pattern of some elderly bearing in mind the findings of Berger et al.[11].In people. However, the nursing home person is more this study, 15 elderly people with type 2 diabetes and likely to suffer from episodes of non-ketotic hyperos- a fasting blood glucose (FBG) level of >9 mmol l−1 molar coma, often triggered by an infection such as a all claimed to feel well and to be asymptomatic, but foot infection or urinary tract infection. Thus, the judi- agreed to start insulin therapy. Over the 8 months of cious use of insulin therapy in elderly persons at risk of these complications is warranted. As a consequence the study period there was a significant fall in HbA1c, and at the end of the study only two people wished of this situation, individuals living in nursing homes to return to tablet therapy. The authors concluded that and residential care require a structured care plan with ‘asymptomatic’ hyperglycaemia in elderly people is regular reviews of both the goals of treatment and the educational needs of the formal carers. often associated with a reduced well being which can only be unmasked by a trial of insulin. The elderly are more likely to have medical comor- 18.4 Some of special considerations bidities, such as renal disease and heart failure, which of elderly people and insulin may preclude the use of other glucose-lowering med- therapy ications. Metformin is contraindicated in persons with compromised renal function [12], and it is generally Cognitive dysfunction is common in the elderly, and recommended that the administration of this agent is individuals with diabetes are not exempt from this. ceased in those individuals with an estimated glomeru- −1 Whilst some studies have demonstrated the level of lar filtration rate (GFR) of <40 ml min . A study by dysfunction correlates with degree of hyperglycaemia Beebe and Patel [13] showed that the glitazones func- [14] the correction of hyperglycaemia generally only tioned well in the elderly, although significant fluid has, at most, marginal effects on improving the situa- retention with a worsening of previously compensated tion. So, the assessment of mental capacity is not so or unrecognized heart failure is a well-known problem much to determine if insulin is indicated but rather to in this age group. Moreover, as a general principle, evaluate to what extent a person can cope with insulin considering the recent information regarding the use therapy without assistance. of glitazones in those with ischaemic heart disease, Elderly people are often more frail, and an assess- the glitazones probably should be used with caution, ment of their physical function is also essential. Whilst and not as first line agents in the elderly. The gli- most current insulin delivery devices appear simple to tazones have also been shown to increase peripheral use, they still require a certain level of manual dex- fractures and to reduce bone density, and for this reason terity. Nevertheless, studies have generally found that are also not preferred agents in the elderly. In elderly prefilled insulin pens are well accepted and used effec- patients the potent long-acting sulphonylurea gliben- tively by elderly patients [15]. An assessment of vision clamide has the potential to cause hypoglycaemia is also required to ensure that an elderly person can which is severe, persistent, resistant to treatment, and administer the correct insulin dosage. About one-third not infrequently results in death or permanent disabil- of elderly people with diabetes have some degree of ity. Among this group of susceptible elderly patients, visual disability, often due to cataract, glaucoma or insulin represents a safer therapeutic alternative than macular degeneration rather than to retinopathy [16]. glibenclamide. Hence, the use of a device that ‘clicks’, so that the 268 CH 18 INSULIN THERAPY number of units of insulin dialled up for delivery can The question of whether combined oral drug and in- be heard and counted, would be necessary in some sulin treatment provides a better glycaemic control than cases. insulin alone has been examined extensively. Yet, this Before commencing an older person on insulin ther- is in a sense a meaningless question, because the an- apy, apart from a mental and physical assessment (as swer would depend on how much insulin was used. described above), a thorough history of their social sit- Although the oral agents have ‘failed’ in the situation uation should also be taken. Typical questions asked of ‘secondary failure’, they are still exerting consider- would include “Who lives at home with you?”; “Do able effects in lowering glucose levels. Clinical studies you cook for yourself?”; and “Does someone visit you have shown that if either sulphonylurea or metformin regularly?”. The replies will help to provide impor- is stopped, then each would need to be replaced by an tant information about the particular individual, such extra 20–30 units of insulin per day. In other words, if as how great is the risk of hypoglycaemia, what is the sulphonylurea and metformin at maximum dosage are best time for the insulin to be administered and, indeed, stopped, the insulin dosage would generally need to ex- if insulin is needed at all. ceed about 40–60 units per day before any significant Often, it is the clinician who feels that the elderly improvement in glycaemic control would occur. All person may have a fear of injections, and so is reluctant too commonly, a deterioration in glycaemic control has to broach the subject. However this is rarely the case been witnessed when both oral drugs were stopped and if the situation is handled sensitively and not used as not replaced with sufficient insulin. An insulin-only a threat. All people with type 2 diabetes should be therapy would require a more complicated regimen and advised early during the course of their disease that the a more rapid titration schedule, making the process of natural history of diabetes is such that approximately beginning insulin treatment much more difficult, as it 50% will require insulin after about 5 years [17, 18]. would involve more clinic visits and a higher risk of In this way, they will realize that the need to receive hypoglycaemia. Whilst applicable to all age groups, insulin therapy is often inevitable, and is not their this would create particular difficulties for the elderly. ‘fault’. It is often also helpful to ask the patient if they It is, however, quite common to modify the oral know anyone else who takes insulin therapy, as this can agent regimen in the following situations: influence their acceptance or, not, of such treatment. • If the person is taking a supra-maximal dosage of any It is also a good opportunity to dispel the myths and oral drugs, it is reduced to what is recommended in misconceptions related to insulin therapy; an example the product information. is that “...my father had an amputation after he went • onto insulin”. If the person is suffering from the gastrointestinal One practical tip used in the authors’ Diabetes Cen- adverse effects of metformin, it is reduced to a tre is that, whenever the subject of insulin therapy is dosage which is tolerated. first raised, the person is assisted in giving themselves • If the person is taking a third glucose-lowering a practice injection. In this way they realize the needle medication, such as acarbose or a glitazone, this does not hurt, the insulin is simple to administer, and would be stopped. In the case of acarbose, this would they leave the consultation feeling that they are capable be because of the drug’s relatively weak action and of commencing such therapy. Having their anxieties al- inconvenient side effect of flatulence. In the case layed also means that a patient will be more likely to of a glitazone, it would be because of potentially accept insulin therapy! serious side effects in the face of a presumed lack of great efficacy (otherwise, insulin therapy would not 18.5 Glucose-lowering medications be required). and the introduction of insulin • If the person chooses to have more insulin injections therapy and at a higher dosage (which, in our experience, is a rare situation). Today, it is standard practice – not only in the elderly Generally, it is easier to persuade patients to undertake but virtually for all patients – to start insulin therapy combined oral drugs and insulin treatment. They are but to maintain the same oral glucose-lowering med- often comforted by the knowledge that this allows them ications, at least for the first year of insulin therapy. to take insulin only once a day, in the privacy of their 18.6 WHICH INSULIN REGIMEN SHOULD BE USED IN THE ELDERLY? 269 own home, and without a great deal of disturbance to With regards to the choice of insulin, until recently their daytime routine. It is also a gentle way of easing an isophane type of insulin has normally been used. them into insulin therapy. When they are familiar with The availability of long-acting insulin analogues such insulin injections they will be more accepting if a as Lantus and Levemir has today made them quite second injection were to be required. popular choices [19], mainly because they have the Occasionally, patients develop frequent daytime hy- advantage of longer and flatter actions. Consequently, poglycaemia on combined treatment. When this hap- they are less likely to cause hypoglycaemia and more pens, the sulphonylurea dosage should be reduced or likely to require only one injection a day. Their phar- ceased if necessary. Apart from this, and in the ab- macokinetic properties also make less important the sence of any contraindications (e.g. renal failure or timing of their injection and relationship to meal times. allergy), there is no good evidence to indicate that For example, if an elderly person needs to have insulin glucose-lowering medications must be stopped at any administered by a relative or a nurse who can only visit stage. In fact, the current policy is to continue them at mid-morning tea, it is perfectly safe for the patient to while the glycaemic control remains satisfactory, albeit maintain their existing meal schedule and to have the at a reduced dosage if the patient has renal or cardiac insulin injected before or after morning tea. Provided comorbidities. Some patients may wish to reduce the that the injections are given at about the same time number of tablets they take each day, especially if they each day, the precise relationship to meals is not of are receiving multiple medications for blood pressure great concern. On this basis, if cost is not a consider- and lipid control. There is no problem in reducing one ation, the use of a long-acting insulin analogue would or more of the oral hypoglycaemic drugs when the pa- be recommended over a conventional isophane insulin. tient has been established on insulin therapy, but it is It is important that the premixed insulins should not be essential that the insulin dose is titrated. administered just before bed time, as their short-acting component would most likely cause nocturnal hypo- 18.6 Which insulin regimen should glycaemia. On the other hand, the main food intake be used in the elderly? of some patients – including some elderly people – is a major meal in the evening. This may be a lifetime habit which is difficult to break, or it may be neces- Bearing in mind the above considerations, the most sary because this is the only time that family members commonly used insulin regimen is to introduce a can gather for a meal. In this situation, there is of- long-acting insulin before bedtime, while maintaining ten a large postprandial hyperglycaemic peak in the oral agent treatment. The starting dose should be low; evening. Under such circumstances, a premixed insulin 10 units would normally be a reasonable starting point. As insulin is an ‘additional’ treatment, the patient’s given before dinner is a good alternative of providing condition will not deteriorate and usually will start to both post-prandial and basal insulin supplies [20]. For improve from that point onward. If the patient is very this, the type of premixed insulin which contains a nervous or reluctant, and it is imperative to minimize short-acting insulin analogue is favoured, because its the risk of hypoglycaemia – however small – then an rapid onset and shorter action profile is best for con- even lower insulin dosage can be used to start the pro- trolling post-prandial hyperglycaemia, with less risk of cess and to gain the patient’s confidence. causing nocturnal hypoglycaemia.

Set individual goals Assess ability to self Assess well being, and metabolic targets inject, cognition and HbA1c +/- symptoms and determine insulin social situation regimen

Review–ensure no Commence insulin hypoglycaemia and therapy – titration by person feels well and elderly person or is coping with health professional treatment 270 CH 18 INSULIN THERAPY 18.7 Titrating the insulin dosage and provided when starting insulin may confuse elderly patients – and, for that matter, many young patients! monitoring progress If the patients are overloaded with information on a stressful day such as this, they may not remem- A major feature of combination therapy is that insulin ber the more important messages, and indeed some is added to an existing treatment. Glycaemic control may even be scared away from insulin treatment al- should therefore improve immediately and, for practi- together. The best practice is to concentrate on teach- cal purposes, should not deteriorate. This means that ing the patient or carer how, and when, to inject the the dose of insulin can be increased slowly, minimizing insulin subcutaneously into the abdomen, using injec- the risk of hypoglycaemia. A titration regimen previ- tion devices which are extremely user-friendly, such ously described suggests increasing the insulin dosage that the technique can be taught in a matter of min- −1 by 4 units per day if the FBG exceeded 8 mmol l utes. The day of starting insulin therapy is also not on three consecutive days, and by 2 units per day if it an ideal time for detailed dietary advice. The cur- −1 exceeded 6 mmol l . Depending on the patients, this rent practice is to emphasize only the need to have can sometimes be implemented directly by the patients regular meals and snacks (including one before bed) or assisted by health professionals after discussing the containing carbohydrates. In fact, the patients are often blood glucose-monitoring results by telephone. In the told that “nothing else needs to change”, and expe- elderly, such a titration can be made slightly more rience suggests that both patients and relatives de- slowly, such as every 1–2 weeks. This slower pace will rive considerable comfort and reassurance from this help to gain the person’s confidence and reduce the risk approach. of hypoglycaemia. In accordance with the philosophy At this stage of diabetes, most patients would be that, in elderly people, it is not always necessary – or familiar with glucose monitoring and should be asked indeed desirable – to aim for excellent glycaemic con- to perform the task [21]. As the adjustment of insulin trol, a less ambitious FBG target is sometimes used. dosage in this regimen is primarily dependent on the For patients in their seventh decade of life, the aim morning FBG concentrations, testing at this time point is to achieve a FBG of 7 mmol l−1, while for those is the first priority and should be included every day. in their eighth decade the target is 8 mmol l−1,andso For some elderly patients who cannot test their blood forth. glucose, it may be necessary to commence insulin with- After 2–3 months, the patient is likely to be receiv- out such monitoring, and to rely on blood glucose ing about 20–30 units of insulin each day, together monitoring at the doctor’s surgery, at the same time with maximum oral drug therapy. Measurement of the measuring HbA1c concentrations and assessing symp- HbA1c concentration after this interval helps to quan- toms to make dose adjustments. tify the overall level of glycaemic control, such that Hypoglycaemia is the only risk in starting insulin further increases in insulin dosage can be made accord- therapy, and however much such a risk is minimized, ingly. There is often a reduction in HbA1c of up to 2% it cannot be completely eliminated [22]. How much in- and an increase in body weight of several kilograms. formation should be provided to patients – especially If these changes are not evident, then the possibility the elderly ones – is a difficult question. Too much should be considered that the patient has not been tak- detail would incur the risk of scaring a reluctant pa- ing the insulin regularly, or that somebody unfamiliar tient away from the correct treatment, but not enough with the regimen has either reduced or stopped one or information would open the door to negligence. This more of the oral hypoglycaemic drugs. At any time, dilemma is, of course, not unique to commencing in- if a patient is noticing significant hypoglycaemia, the sulin, and each doctor must make a decision with an sulphonylurea and then the insulin dosage should be individual patient. It is reassuring that, in patients with reduced. type 2 diabetes, hypoglycaemia due to insulin is usually less severe. When possible, the patient should also be seen a few 18.8 What should patients be told days after starting insulin therapy in order to review on the day they start insulin? their injection technique. If this is satisfactory, then the stabilization of glycaemic control can be achieved Although everyone has different educational needs by telephone or e-mail communication every few days, and capacities to learn, the comprehensive information until the target blood glucose levels are achieved. In 18.10 FAVOURABLE OUTCOMES 271

Table 18.1 Clinical data acquired from patients with type 2 diabetes (n = 493)who attended the authors’ Diabetes Centre, and were commenced on insulin therapy. Parameter Age of patients

<60 years (n = 218) ≥60 years (n = 275) Duration of diabetes at start of insulin 8.1 (6.3–11.4) 12.0 (8.2–16.2) therapy (years) Insulin dosage on once-daily injection 23 (20–33) 20 (14–25) (units) Change in HbA1c (%) from start of −0.3 (−1.4−0.6) −0.4(−1.0−0.3) insulin therapy to first review Time between once- and twice- daily 2.0 (1.2–3.8) 2.2 (1.3–3.7) injections (years) No. of injections (%) 16159 23136 ≥385 Hypoglycaemia reported at last visit 1.8 2.9 (%) this situation a family member or friend is invaluable the present authors’ Diabetes Centre were commenced in assisting with communication, and it is normally on insulin therapy (see Table 18.1). Of these, 275 suggested that such a companion attend at least the individuals were aged >60 years. first visit, if at all possible. The goal of treatment in elderly patients is to optimize glycaemic control to an extent consistent 18.9 When should more complex with the avoidance of hypoglycaemia and the prevention of acute complications. Age-related physiological insulin regimens be changes, life expectancy and comorbidities will influ- introduced? ence the type of treatment implemented. Consequently, the management plan should be individualized so as After a couple of years on a single insulin injection, to ensure that the elderly person enjoys a balance be- a second injection – usually given in the morning between quality of life and the diabetes regimen. In this, fore breakfast – may need to be introduced, depending the ‘art’ of medicine will prevail, even more than the on the subject’s age and well-being. In this situation, ‘science’ of medicine. a small starting dose of medium-acting insulin or a premixed insulin in the order of 6–12 units would be reasonable [23]. Based on experience, patients nor- References mally accept additional insulin injections much more readily than the first one, having realized that the in- 1. DeFronzo, R.A. (1988) Pathogenesis of type 2 dia- jection is not a major difficulty. Occasionally, elderly betes mellitus. Medical Clinics of North America, 88, individuals may adapt quite well to use of a basal bo- 787–835. lus regimen, and in some cases also to insulin pump 2. Gossain, V.V., Carella, M.J. and Rovner, D.R. (1994) therapy [24]. Management of diabetes in the elderly: a clinical perspective. Journal of the Association for Academic Mi- nority Physicians, 5, 22–31. 18.10 Favourable outcomes 3. Gambert, S.R. (1990) Atypical presentation of diabetes mellitus in the elderly. Clinics in Geriatric Medicine, According to a computerized database, elderly people 6, 721–9. respond to insulin therapy in a similar manner to 4. Rosenstock, J. (2001a) Insulin therapy: optimizing con- their younger counterparts. Between 2003 and 2006, a trol in type 1 and type 2 diabetes. Clinical Cornerstone, total of 493 individuals with type 2 diabetes attending 4, 50–64. 272 CH 18 INSULIN THERAPY

5. Bourdel-Marchasson, I., Dubroca, B., Manciet, G., De- 15. Coscelli, C., Lostia, S., Lunetta, M., Nosari, I. and camps, A., Emeriau, J. P. and Dartigues, J.F. (1997) Coronel, G. A. (1995) Safety, efficacy, acceptability Prevalence of diabetes and effect on quality of life in of a pre-filled insulin pen in diabetic patients over 60 older French living in the community: the PAQUID years old. Diabetes Research & Clinical Practice, 28, Epidemiological Survey. Journal of the American Geri- 173–7. atrics Society, 45, 295–301. 16. Klein, R. (1991) Age-related eye disease, visual im- 6. Hendra, T.J. (2002) Starting insulin therapy in elderly pairment, and driving in the elderly. Human Factors, patients. Journal of the Royal Society of Medicine, 95, 33, 521–5. 453–5. 17. Matthews, D.R., Cull, C.A., Stratton, I.M., Holman, 7. Gale, E.A., Dornan, T.L. and Tattersall, R.B. (1981) R.R. and Turner, R.C. (1998) UKPDS 26: Sulpho- Severely uncontrolled diabetes in the over-fifties. Dia- nylurea failure in non-insulin-dependent diabetic betologia, 21, 25–8. patients over six years. UK Prospective Diabetes 8. Brown, A.F., Mangione, C.M., Saliba, D., Sarkisian, Study (UKPDS) Group. Diabetic Medicine, 15, C.A. and the California Healthcare Founda- 297–303. tion/American Geriatrics Society Panel on Improving 18. Sinclair, A.J. and Meneilly, G.S. (2000) Re-thinking Care for Elders with diabetes. (2003) Guidelines for metabolic strategies for older people with type 2 di- improving the care of the older person with diabetes abetes mellitus: implications of the UK Prospective mellitus. Journal of the American Geriatrics Society, Diabetes Study and other recent studies. Age & Ageing, 51, S265–80. 29, 393–7. 9. Meneilly, G.S. and Tessier, D. (2001) Diabetes in 19. Rosenstock, J., Schwartz, S.L., Clark, C.M., Jr, Park, elderly adults. [see comment]. Journals of Gerontology G.D., Donley, D.W. and Edwards, M.B. (2001) Basal Series A – Biological Sciences and Medical Sciences, insulin therapy in type 2 diabetes: 28-week compari- 56, M5–13. son of insulin glargine (HOE 901) and NPH insulin. 10. Singh, I. and Marshall, M.C., Jr (1995) Diabetes melli- Diabetes Care, 24, 631–6. tus in the elderly. Endocrinology and Metabolism Clin- 20. Coscelli, C., Calabrese, G., Fedele, D., Pisu, E., ics of North America, 24, 255–72. Calderini, C., Bistoni, S., Lapolla, A., Mauri, M.G., 11. Berger W. (1988) Insulin therapy in the elderly type Rossi, A. and Zappella, A. (1992) Use of premixed 2 diabetic patient. Diabetes Res Clin Pract 4 (Suppl), insulin among the elderly. Reduction of errors in 124–8. patient preparation of mixtures. Diabetes Care, 15, 12. Cusi, K., Consoli, A. and Defronzo, R. A. (1996) 1628–30. Metabolic effects of metformin on glucose and lactate 21. Skyler, J.S. (1997) Glucose control in type 2 diabetes metabolism in noninsulin-dependent diabetes mellitus. mellitus. Annals of Internal Medicine, 127, 837–9. Journal of Clinical Endocrinology and Metabolism, 81, 22. Rosenstock, J. (2001) Management of type 2 diabetes 4059–67. mellitus in the elderly: special considerations. Drugs & 13. Beebe, K. and Patel, J. (2003) Rosiglitazone is effective Aging, 18, 31–44. and well tolerated in patients >65 years with type 2 23. Turner, H.E. and Matthews, D.R. (2000) The use of diabetes [abstract]. Diabetes 99; 42 (Suppl. 1), A111. fixed-mixture insulins in clinical practice. European 14. Reaven GM, Thompson LW, Nahum D and Haskins Journal of Clinical Pharmacology, 56, 19–25. E. (1990) Relationship between hyperglycemia and 24. Rizvi AA. (2002) Benefits of insulin pump therapy in cognitive function in older NIDDM patients. Diabetes the elderly. Geriatric Times 3 (4), 23–30. Care, 13(1), 16–21. 19 Treatment of Hypertension

Peter Fasching 5th Medical Department, Wilhelminenspital, Vienna, Austria

Key Messages 19.2 Risk assessment from observational studies • Hypertension is the commonest complication associated with diabetes mellitus. A large pro- In a meta-analysis of 61 prospective observational portion of older patients may have unsatisfactory studies of blood pressure and mortality which in- blood pressure levels, despite treatment. cluded 9 58 074 participants and an analysis of 12.7 • Treatment of hypertension can be effective in million person-years at risk, it was shown that within substantially reducing cardiovascular risk, stroke each decade of age a proportional beneﬁt of a low rate and death. • A general blood pressure target is 140/80 mmHg blood pressure down to 115 mmHg systolic blood or less in patients with diabetes, but much lower pressure (SBP) and 75 mmHg diastolic blood pres- levels should be aimed for in those able to sure (DBP), was evident [3]. At the age of 40–69 tolerate therapy. years, each reduction of 20 mmHg SBP (or, approx- • Treatment of hypertension in patients with imately equivalently, of 10 mmHg usual DBP) is as- diabetes should include an ACE inhibitor or sociated with more than a twofold lowering of the angiotensin-receptor blocker. death rate from stroke, and with a twofold lowering of death rates from ischaemic heart disease and other vascular causes. Whilst all of these proportional re- 19.1 Introduction ductions in vascular mortality are only about a half in persons aged 80–89 years when compared with According to the results of recently published studies, subjects aged 40–49 years, the annual absolute dif- hypertension (with a prevalence of 66%) is the most ference in risk is greater in old age. For predicting frequent comorbidity in type-2 diabetic patients aged vascular mortality from a single blood pressure mea- >65 years [1]. Adequate diabetic management and pre- surement, the average of SBP and DBP is slightly vention of cardiovascular complications are regarded more informative than either alone, and pulse pres- as crucial for controlling health care costs in this group sure is much less informative. Yet, the number of of patients. On the other hand, a recent survey showed included diabetic patients is not speciﬁed in the re- that 73% of male diabetic patients with a mean age of port. In contrast, smaller epidemiological studies have 66 years had a blood pressure >140/90 mmHg and re- shown that in men aged ≥85 years, a higher sys- ceived less intensive antihypertensive medication ther- tolic blood pressure is associated with a better survival apy than patients without diabetes [2]. [4, 5].

Table 19.1 Indications for initial treatment and goals for adult hypertensive diabetic patients, according to ADA (2008). Systolic BP (mmHg) Diastolic BP (mmHg) Goal (mmHg) <130 <80 Lifestyle therapy alone (max. 3 months), then 130–139 80–89 add pharmacological treatment Lifestyle therapy + pharmacologic treatment >140 >90

19.3 Target values for hypertension In the recent European guidelines [7], Grade 1 hypertension (mild) is defined as SBP 140–159 mmHg treatment in patients with and/or DBP 90–99 mmHg. An isolated systolic hyper- diabetes tension (ISH) is present if the SBP is >140 mmHg and DBP is <90 mmHg. In patients aged over 50 years, a Several authorities have issued guidelines for treatment SBP >140 mmHg represents a markedly more impor- of blood pressure within diabetic populations. The tant risk factor for cardiovascular disease (CVD) than American Diabetes Association recommendations are does the DBP. listed in Table 19.1, whilst other published recommen- As a therapeutic aim, a reduction in SBP and dations are listed in Tables 19.2–19.4, for comparison. DBP below 140/90 mmHg is recommended in all pa- In the American JNC 7 Report [6], a SBP of tients who tolerate this, and in diabetic patients be- 120–139 mmHg or a DBP of 80–89 mmHg is defined low 130/80 mmHg. Yet, it must be recognized that in as ‘Pre-hypertension’, and lifestyle modification is ad- ISH in older patients it is difficult to reach systolic vised to prevent cardiovascular complications. Riva-Rocchi (RR)-values below 140 mmHg. Table 19.2 Blood pressure control assessment level All recommendations have in common that they do according to diabetes type 2 Desktop Guide- not differentiate with regards to the age of the pa- lines 1998–1999. European Diabetes Policy Group tients. The recommendation by the ADA [9] sets a (IDF-European Region) [8]. goal of <130/<80 mmHg systolic/diastolic, which is most ambitious compared to the older IDF or recent • Low risk (mmHg) <140/85 German guidelines [10] (where the common goal • Life-style management of raised blood pressure should is <140/<85 mmHg). A European-wide primary care be given a good trial before beginning antihypertensive approach to the treatment of hypertension in diabetes drugs is provided in Table 19.4. • If antihypertensive drugs are necessary use: None of the previous guidelines has included spe- ◦ single-agent therapy at rising doses until target cific and detailed sections on managing the older pa- achieved (or intolerance) tient, or those with frailty. Evidenced-based clinical ◦ multiple therapy if targets not reached on maximum guidelines for type 2 diabetes in older patients were doses of single agents presented in 2004 [12], and recommendations were ◦ once-daily drug administration regimes made in these areas. A more recent update is planned

Table 19.3 Therapeutic target values for antihypertensive therapy for adult patients with diabetes mellitus type 2, according to the clinical guidelines of the German Diabetes Association (Praxis-Leitlinien der Deutschen Diabetes Gesellschaft (DDG). Clinical scenario Therapeutic target for blood pressure Systolic BP (mmHg) Diastolic BP (mmHg) Diabetes with essential hypertension <140 <85 Good tolerance of BP <140/85 mmHg <130 <80 Diabetes with microalbuminuria <130 <80 With/or manifest nephropathy better <120 – 19.4 AVAILABLE EVIDENCE FROM RANDOMIZED CONTROLLED TRIALS 275

Table 19.4 Recommendations according to the SVDPCDG Guidelines Working Group. St Vincent Declaration Primary Care Diabetes Group. Approved Guidelines for Type 2 Diabetes 2002 [11]. • The general practitioner (GP) determines body mass index (BMI) based on height and weight and blood pressure. A patient with DBP >85 mmHg or SBP >150 mmHg is a candidate for follow-up measurements. • Determination of the urine albumin concentration or the albumin:creatinine ratio in patients aged >50 years is not recommended. From a cardiovascular point of view, this group should already be receiving optimal treatment by a tight regulation of dyslipidaemia and reduction of blood pressure to <150/85 mmHg. It is not known whether the treatment of microalbuminuria in patients aged >50 years, without hypertension, is worthwhile in the prevention of nephropathy. • There is evidence that, in patients with type 2 diabetes mellitus, maintaining the DBP <90 mmHg and SBP <160 mmHg reduces the risk of microvascular and macrovascular complications. It is, therefore, recommended that the target blood pressure should be <150/85 mm Hg. Whether the reduction of blood pressure is achieved using thiazide diuretics, ACE-inhibitors or β-blockers leads to virtually the same reduction in microvascular and macrovascular complications. • The following steps are recommended: ◦ Step 1: a low dose of thiazide diuretic, hydrochlorothiazide or chlorthalidone 12.5 mg ◦ Step 2: add an ACE-inhibitor or a β-blocker. ◦ Step 3: a combination of the three aforementioned drugs. • Any initial kidney function deterioration must be monitored when prescribing an ACE-inhibitor in older patients.

for the IDF Conference in Montreal, 2009, when the against the risk of myocardial infarction (MI) [14]. threshold for treatment is expected to recommended A meta-regression across 27 trials (1 36 124 patients) as 140/80 mmHg or higher present for 3 months, and showed that the odds ratio could be explained by measured on at least three separate occasions. A lower the achieved differences in systolic pressure. Thus, value is recommended for those patients who are able these findings emphasize that blood pressure control to tolerate the therapy and to self-manage, and/or those is important. Although all antihypertensive drugs have with concomitant renal disease. For frail subjects (de- similar long-term efficacy and safety, calcium chan- pendent; multisystem disease; care home residency, nel blockers might be especially effective in stroke including those with dementia), an acceptable blood prevention. It was not found that ACE inhibitors or pressure is recommended to be <150/90 mmHg. Al- alpha-receptor blockers affected cardiovascular prog- though this is supported in part by others, a lower DBP nosis beyond their antihypertensive effects [14]. is suggested for the oldest of elderly subjects (aged Similar conclusions have also been reached follow- >80 years), namely <150/80 mmHg [13]. ing an overview of placebo-controlled trials of ACE inhibitors, calcium channel blockers and other blood pressure-lowering drugs (15 studies; 74 696 individ- 19.4 Available evidence from uals; mean age 62 years; no specification as to age randomized controlled trials or presence of diabetes, although studies as ABCD, HOPE and Syst-Eur have been included) [15]. 19.4.1 Meta-analysis: old versus new drugs Recently, it was questioned whether beta-blocker therapy should remain first choice in the treatment The first meta-analysis (which included nine random- of primary hypertension, since in a meta-analysis in ized trials comparing treatment in 62 605 hypertensive comparison with other antihypertensive drugs the ef- patients) investigated whether antihypertensive drugs fect appeared to be less than optimum with a raised offered cardiovascular protection beyond blood pres- risk of stroke [16]. A Cochrane analysis also con- sure lowering. Compared with ‘old’ drugs (e.g. di- cluded that the available evidence does not support uretics and β-blockers), calcium-channel blockers and the use of beta-blockers as first-line drugs in the treat- angiotensin-converting enzyme (ACE) inhibitors of- ment of hypertension. This conclusion was based on fered a similar overall cardiovascular protection, but the relatively weak effect of beta-blockers to reduce the calcium-channel blockers provided more reduc- stroke and the absence of an effect on coronary heart tion against the risk of stroke but less reduction disease when compared to placebo or no treatment. 276 CH 19 TREATMENT OF HYPERTENSION More importantly, it was based on the trend towards In fact, chlorthalidone was superior to lisinopril for worse outcomes in comparison with calcium-channel several CVD outcomes, and superior to amlodipine blockers, renin–angiotensin system (RAS) inhibitors for heart failure in both the diabetic and non-diabetic and thiazide diuretics. Most of the evidence of these participants. The study arm with doxazosin was conclusions was derived from trials where atenolol was stopped at an early stage because of a large excess used as the beta-blocker (75% of beta-blocker par- of heart failure with a need for hospitalization. ticipants in this review). However, it is not known Doxazosin should therefore be avoided as a first-line at present whether beta-blockers have differential ef- antihypertensive drug, and particularly so in patients fects on younger and elderly patients, or whether with heart failure [21]. there are differences between the different subtypes of One detrimental effect of diuretic therapy in the beta-blocker [17]. ALLHAT study was a slightly elevated fasting blood With regards to the age-dependent effect, another glucose in non-diabetic patients compared to the other meta-analysis advised that beta-blockers should not treatment options. be considered first-line therapy for older hypertensive Especially, high-dose diuretic therapy and patients without another indication for these agents. β-blockers can cause a significant increase in the However, in younger patients beta-blockers were as- prospective diabetic risk [22]. Of interest also was a sociated with a significant reduction in cardiovascular recent post-hoc analysis of the Systolic Hypertension morbidity and mortality [18]. In comparison to other in the Elderly Program (SHEP), which showed that agents, a systematic review by Bradley et al. found the participants who had hypokalaemia after one beta-blockers to be inferior to calcium channel blockers year of treatment with a low-dose diuretic did not and also to inhibitors of the RAS for reducing sev- experience the reduction in cardiovascular events eral important hard endpoints. Compared to diuretics, achieved among those without hypokalaemia [23]. they had similar outcomes, but were less well toler- The HOPE study, where the active treatment ated [19]. ACE-inhibitor ramipril (10 mg) was compared to placebo, showed cardiovascular protection in a 19.4.2 Prospective studies: old versus large group of individuals with high cardiovascular new drugs risk, including also diabetic persons [24]. In the The Antihypertensive and Lipid-Lowering Treatment OnTarget study, the angiotensin-receptor blocker to Prevent Heart Attack Trial (ALLHAT) was a (ARB) telmisartan (80 mg per day) was compared placebo-controlled trial conducted to determine to the ‘gold standard’ ramipril (10 mg per day), and whether treatment with a calcium-channel blocker to a combination of both drugs, in over 25 000 or an ACE inhibitor would lower the incidence of persons with high cardiovascular risk. After a mean coronary heart disease (CHD) or other CVD events in follow-up of 56 months, telmisartan was equivalent to comparison to treatment with a diuretic [20]. After a ramipril in patients with vascular disease or high-risk mean follow-up of 4.9 years in a group of 33 357 par- diabetes, and associated with less angioedema. The ticipants aged ≥55 years (mean age 66.9 years) who combination of the two drugs was associated with a had at least one other CHD risk factor, thiazide-type greater number of adverse events, but without any diuretics (chlorthalidone, 12.5–25 mg daily) were seen increase in benefit [25]. At baseline, the mean age to be equal in preventing one or more major forms of patients was 66 years, and 38% had diabetes. In a of CVD, and to be less expensive. Approximately subgroup analysis, age (also for those aged >75 years) 36% of the participants were known diabetics at the and the presence of diabetes showed no influence on start of the study. The results, even when comparing the study results. individual therapy arms, were also consistent for 19.4.3 Prospective studies in old people subgroups of participants – that is, older and younger, male and female, diabetic and non-diabetic. Thus, for A subgroup meta-analysis of randomized, controlled the important diabetic population, lisinopril appeared trials of antihypertensive drugs in very old people to have no special advantage, while amlodipine had (aged >80 years; n = 1670) suggested that treatment no particular detrimental effect for most CVD and prevented 34% of strokes. The rates of major car- renal outcomes when compared to chlorthalidone. diovascular events and heart failure were significantly 19.4 AVAILABLE EVIDENCE FROM RANDOMIZED CONTROLLED TRIALS 277 decreased, by 22% and 39%, respectively. However, and hinted at a degree of selection bias towards very there was no treatment benefit for cardiovascular death, healthy old individuals. and a significant 6% (range from – 5% to 18%) relative excess of death from all causes. These inconclusive 19.4.4 Prospective studies in type 2 findings for mortality contrasted with the benefit of diabetic patients treatment for non-fatal events. The percentage of diabetic patients was reported as 14% of the total number. A total of 11 140 type 2 diabetic patients (mean age No separate analysis was carried out, however, most 66 years) was enrolled in the ADVANCE trial, con- likely due to the small absolute number [26]. ducted by the Advance Collaborative Group [29]. The In the STOP-Hypertension-2 trial, a total of patients were randomized to a fixed combination of the 6614 patients aged 70–84 years with hypertension ACE-inhibitor perindopril and indapamide or placebo, (RRsystolic >180 mmHg; RRdiastolic >105 mmHg, or in addition to current therapy. After a mean of 4.3 both) were randomly assigned to conventional anti- years of follow-up, those assigned active therapy had hypertensive drugs (β-blockers, diuretics) or newer a mean reduction in SBP of 5.6 mmHg and in DBP of drugs (ACE-inhibitors, calcium-channel blockers) 2.2 mmHg. The relative risk of a major macrovascular [27]. The blood pressure was seen to decrease to a or microvascular event was reduced by 9% (p = 0.04), similar degree in all treatment groups, and the ‘old’ of death from CVD by 18% (p = 0.03), and of death and ‘new’ antihypertensive agents were similar in from any cause by 14% (p = 0.03). Hence, in this terms of their prevention of cardiovascular mortality study, over a period of 5 years, the number of patients or major events. A decrease in blood pressure was of required to be treated in order to prevent one death was major importance for the prevention of cardiovascular 79 for the blood pressure arm. events. Among the total patient cohort, 10.9% Major benefits of blood pressure-lowering in (n = 719; 433 females, 286 males, mean age 75.8 type 2 diabetic patients by treatment with either years) had diabetes at baseline, but the treatment an ACE-inhibitor or β-blocker versus placebo effects did not differ significantly in terms of the were demonstrated by the hypertension arm of the primary endpoint (treatment duration 4–5 years). UKPDS some 10 years earlier [30, 31], albeit with In a subgroup analysis of the Syst-Eur Study [28], a higher median blood pressure in each group. The diabetic patients aged >70 years showed a greater ADVANCE blood pressure values at baseline were clinical benefit of the blood pressure-lowering afforded similar to those in the tightly controlled groups of the by the calcium channel-blocker nitrendipine than their UKPDS. age-matched, non-diabetic counterparts. The long-term follow-up data acquired for the tight In the very recently reported HYVET study [13], a control of type 2 diabetes was published some 10 years total of 3845 patients aged >80 years (mean age 83.6 after that of the UKPDS, in 1998 [32]. The differ- years) with a sustained SBP ≥160 mmHg received (at ences in blood pressure between treatment groups had random) either active treatment with the diuretic in- disappeared within 2 years following termination of dapamide, or a matching placebo, The ACE-inhibitor the trial. Neither was any risk reduction seen during perindopril or placebo was added if necessary to the long-term follow-up for MI, nor death from any achieve a target blood pressure of 150/80 mmHg. After cause. The suggestion was, therefore, that good blood 2 years the mean blood pressure was 15.0/6.1 mmHg pressure control must be continued if the benefits are lower in the indapamide-treated group than in the to be maintained. Yet, this finding was in contrast to placebo group. This in turn led to a 30% reduction in the 10-year follow-up of intensive glucose control arm the rate of fatal and non-fatal stroke (borderline signif- of the UKPDS, which had been published simulta- icance, p = 0.06) and to a significant (21%) reduction neously [33]. Despite an early loss of the glycaemic in mortality rate from any cause (p = 0.02). Overall, differences between groups, a continued reduction in there was a 39% reduction in death from stroke, and microvascular risk and emergency risk reduction for a 23% reduction in death from cardiovascular causes, MI and death from any cause were observed during after a median follow-up of 1.8 years. Unfortunately, a 10-year post-trial follow-up. A continued benefit af- such a very small proportion (<7%) of diabetic pa- ter metformin therapy was evident among overweight tients at baseline was remarkable in that age group, patients, however. 278 CH 19 TREATMENT OF HYPERTENSION 19.5 Antihypertensive therapy as for patients with diabetes and hypertension should include either an ACE inhibitor or an ARB. If one part of a multifactorial class of drug is not tolerated, then the other should be intervention substituted. If needed to achieve blood pressure targets, a thiazide diuretic should be added to the therapy of The combined effects of glycaemic control and blood those patients with an estimated glomerular filtration pressure control in the UKPDS were reported in an rate (eGFR) >50 ml min−1 per 1.73 m−2, and a loop epidemiological manner, highlighting the additive ef- diuretic for those with an eGFR <50 ml min−1 per fects of both intervention to reduce microvascular and 1.73 m−2. macrovascular complications and the risk of death [34]. Multiple drug therapy (two or more agents at maxi- The recently reported details of the glucose-lowering mal doses) is generally required to achieve blood pres- arm of the ADVANCE study demonstrated a significant sure targets. If ACE inhibitors, ARBs or diuretics are reduction in a combined primary end-point (sum of used, then kidney function and serum potassium levels macro- and microvascular complications). The additive should be closely monitored [9]. efficacy of a tight control of blood glucose and blood In older individuals, kidney function is a pressure in elderly type 2 diabetic patients was demon- critical factor in clinical practice. Severe distur- strated at the EASD meeting in Rome, in September bances of electrolyte concentrations (high and low 2008. Among patients included in the group, with both potassium/hyponatraemia and hypernatraemia are optimal control of blood pressure and blood glucose, common in patients, especially when receiving an 18% reduction in all-cause mortality was observed multiple medications (i.e. ACE-inhibitors, ARBs, (p = 0.04), as well as a 24% reduction in cardiovas- aldosterone-antagonists, diuretics, antidepressants and cular mortality (p = 0.035) and a 33% reduction in antipsychotics inducing the inappropriate secretion of new or worsening nephropathy (p = 0.005). These ef- antidiuretic hormone). Long-acting calcium-channel fects were greater than were observed in each separate blockers are associated with good evidence of treatment arm, while the effects of both interventions stroke prevention, and are neutral with regards to were independent from each other and fully additive. metabolism, electrolytes and kidney function [28]. Thus, antihypertensive therapy in older patients with Beta-blockers should be considered if there is an diabetes must be seen as one part of the multifacto- additive indication besides primary hypertension, as rial intervention necessary in diabetes management to is often the situation in older diabetic patients (i.e. prevent and/or delay cardiovascular complications, as status post-MI, heart failure, tachycardia, coronary shown in the Steno-2 Study [35]. In an observational heart disease). Any contraindications for beta-blocker follow-up of this cohort (n = 160) over another 5.5 therapy must be ruled out before prescribing this class years following the intervention (i.e. a total follow-up (i.e. severe chronic obstructive pulmonary disease, 13.3 years), it appeared that intensive multifactorial acute myocardial insufficiency, critical peripheral dis- therapy was associated with a lower risk of death from ease, bradycardia consequent to sick-sinus syndrome). cardiovascular causes (−57%, p < 0.04), of cardiovas- Diabetes mellitus per se in NOT a contraindication cular events (−59%, p < 0.001), of end-stage renal for beta-blocker therapy. disease (p = 0.04), and also of the need for retinal − = photocoagulation ( 55%, p 0.02) [36]. References The prevention or delay of visual and sensory loss, of coronary, cerebrovascular and peripheral vascular 1. Niefeld MR, Braunstein JB, Wu AW, Saudek CD, events in diabetic patients is of crucial importance in Weller WE and Anderson GF. Preventable hospitaliza- order to preserve self-competence and independence in tion among elderly Medicare beneficiaries with Type 2 further life. diabetes. Diabetes Care 26: 1344–49, 2003. 2. Berlowitz DR, Ash AS, Hickey EC et al. (2003) Hy- 19.6 Pharmacological therapy of pertension management in patients with diabetes: the need for more aggressive therapy. Diabetes Care 26: hypertension in diabetes 355–9. 3. Prospective Studies Collaboration. (2002) Age-specific According to the recommendations of the American relevance of usual blood pressure to vascular mor- Diabetes Association [9], any pharmacological therapy tality: a meta-analysis of individual data for one 19.6 PHARMACOLOGICAL THERAPY OF HYPERTENSION IN DIABETES 279

million adults in 61 prospective studies. Lancet, 360, 16. Lindholm LH, Carlberg B and Samuelsson O. Should 1903–13. beta blockers remain first choice in the treatment of 4. Satish S, Freeman DH, Ray L and Goodwin JS. The primary hypertension. A meta-analysis. Lancet 366: relationship between blood pressure and mortality in 1545–53, 2005. the oldest old. J Am Geriatr Soc 49: 367–374, 2001. 17. Wiysonge CS, Bradley H, Mayosi BM, Maroney R, 5. Van Bemmel T, Gussekloo J, Westemdorp RGJ, Blauw Mbewu A, Opie LH, Volmink J. Beta-blockers for GJ. In a population-based prospective study, no associ- hypertension. Cochrane Database Syst Rev Jan 24 (1): ation between high blood pressure and mortality after CD002003, 2007. age 85years. J Hypertension 24: 287–292, 2006. 18. Khan N and McAllister FA. Re-examining the efficacy 6. Chobanian AV, Bakris GL, Black HR et al.andthe of beta-blockers for the treatment of hypertension: a National High Blood Pressure Education Program Co- meta-analysis. Canadian Medical Association Journal ordinating Committee. The seventh report of the Joint 174: 1737–42, 2006. National Committee on prevention, detection, evalua- 19. Bradley HA, Wiysonge CS, Volmink JA, Mayosi BM tion, and treatment of high blood pressure. The JNC 7 and Opie LH. How strong is the evidence for use Report. JAMA 289: 2560–2572, 2003. of beta-blockers as first-line therapy for hypertension? 7. Guidelines Committee ESH/ESC: 2003 European Soci- Systematic review and meta-analysis. J Hypertens 24: ety of Hypertension – European Society of Cardiology 2131–41, 2006. guidelines for the management of arterial hypertension. 20. The ALLHAT Officers and Coordinators for the J Hypertension 21: 1011–1053, 2003. ALLHAT Collaborative Research Group. Major out- 8. Alberti G., for the European Diabetes Policy Group comes in high-risk hypertensive patients randomised 1998-1999. (1999) Guidelines for Diabetes Care. A to angiotensin-converting enzyme inhibitor or calcium Desktop Guide to Type 2 Diabetes Mellitus. Interna- channel blocker vs. diuretic: the Antihypertensive and tional Diabetes Federation, IDF (European Region). Lipid-Lowering Treatment to Prevent Heart Attack Diabetic Medicine, 16, 716–30. Trial (ALLHAT). JAMA 288: 2981–97, 2002. 9. American Diabetes Association ADA: Standards of 21. The ALLHAT Officers and Coordinators for the ALL- Medical Care in Diabetes - 2008. Hypertension/blood HAT Collaborative Research Group. Major cardiovas- pressure control – Recommendations. Diabetes Care cular events in hypertensive patients randomised to 31 (Suppl. 1): S24–6, 2008. doxazosin vs. chlorthalidon: the Antihypertensive and 10. Scherbaum WA, Landgraf fur¨ die Leitlin- Lipid-Lowering Treatment to Prevent Heart Attack ienkommission der DDG. Praxis-Leitlinien der Trial (ALLHAT). JAMA 283: 1967–75, 2000. Deutschen Diabetes-Gesellschaft (DDG). Diabetes & 22. Gress TW, Nieto J, Shahar E, Wofford MR, Brancati Stoffwechsel, 11 (Suppl. 2): 20, 2002. FL for the Atherosclerosis Risk in Communities Study. 11. Rutten GEHM, Verhoeven S, Heine RJ, De Grauw Hypertension and antihypertensive therapy as risk fac- WJC, Cromme PVM, Reenders K, Van Ballegooie, tors for type 2 diabetes mellitus. N Engl J Med 342: Wiersma TJ and torgnamhe St Vincent Declaration 905–12, 2000. Primary Care Diabetes Group. (1999) Approved Guide- 23. Franse LV, Pahor M, Di Bari M, Somes GW, Cushman lines for Type 2 Diabetes 2002. Huisarts en Wetenschap WC and Apllegate WB. Hypokalemia associated with 42(2): 67–84. diuretic use and cardiovascular events in the Systolic 12. The European Diabetes Working Party for Older Peo- Hypertension in the Elderly Program. Hypertension 35: ple, 2001-2004. Clinical Guidelines on Type 2 Diabetes 1025–30, 2000. Mellitus. Available at www.eugms.org. 24. Heart Outcomes Prevention Evaluation (HOPE) Study 13. Beckett NS, Peters R, Fletcher AE et al.forthe Investigators. Effects of ramipril on cardiovascular out- HYVET Study Group. Treatment of hypertension in comes in people with diabetes mellitus: results of the patients 80years of age or older. N Engl J Med 358: HOPE study and MICRO-HOPE substudy. Lancet 355: 1887–1898, 2008. 253–9, 2000. 14. Staessen JA, Wang J-G and Thijs L. Cardiovas- 25. The ON TARGET Investigators. Telmisartan, ramipril, cular protection and blood pressure reduction: a or both in patients at high risk for vascular events. N meta-analysis. Lancet 358: 1305–15, 2001. Engl J Med 358: 1547–59, 2008. 15. Blood Pressure Lowering Treatment Trialists’ Collab- 26. Gueyffier F, Bulpitt C, Boissel J-P, Schron E, Ekbom oration. Effects of ACE inhibitors, calcium antago- T, Fagard R, Casiglia E, Kerlikowske K, Coope J, for nists, and other blood-pressure-lowering drugs: results the INDANA Group. Antihypertensive drugs in very of prospectively designed overviews of randomised tri- old people: a subgroup meta-analysis of randomised als. Lancet 355: 1955–64, 2000. controlled trials. Lancet 353: 793–6, 1999. 280 CH 19 TREATMENT OF HYPERTENSION

27. Hansson L, Lindholm LH, Ekbom T, Dahlof¨ B, Lanke 32. Holman RR, Paul SK, Bethel MA, Neil HAW, J, Schersten B, Wester P-O, Hedner T, de Faire U, for Matthews DR. Long-term follow-up after tight control the STOP-Hypertension-2 Study Group. Randomised of blood pressure in type 2 diabetes. N Engl J Med trial of old and new antihypertensive drugs in elderly 359: 1565–76, 2008. patients: cardiovascular mortality and morbidity the 33. Holman RR, Paul SK, Bethel MA, Matthews DR and Swedish Trial in Old Patients with Hypertension-2 Neil HAW. 10-year follow-up of intensive glucose study. Lancet 354: 1751–61, 1999. control in type 2 diabetes. N Engl J Med 359: 1577–89, 28. Tuomilehto J, Rastenyte D, Birkenager¨ WH et al.for 2008. the Systolic Hypertension in Europe Trial Investigators. 34. Adler AI, Stratton IM, Neil HAW, Yudkin JS, Effects of calcium-channel blockade in older patients Matthews DR, Cull CA, Wright AD, Turner RC, with diabetes and systolic hypertension. N Engl J Med Holman RR, on behalf of the UK Prospective Diabetes 340: 677–84, 1999. Study Group. Association of systolic blood pressure 29. Advance Collaborative Group. Effects of a ﬁxed com- with macrovascular and microvascular complications bination of perindopril and indapamide on macrovas- of type 2 diabetes (UKPDS 36): prospective cular and microvascular outcomes in patients with type observational study. Br Med J 321: 412–419, diabetes mellitus (the ADVANCE trial): a randomised 2000. controlled trial. Lancet 370: 829–40, 2007. 35. Gaede P, Vedel P, Larsen N, Jensen GVH, Parving 30. UK Prospective Diabetes Study Group. Tight blood HH and Pedersen O. Multifactorial intervention and pressure control and risk of macrovascular and mi- cardiovascular disease in patients with type 2 diabetes. crovascular complications in type 2 diabetes: UKPDS N Engl J Med 348: 383–93, 2003. 38. Br Med J 317: 703–13, 1998. 36. Gaede P, Lund-Andersen H, Parving HH and Pedersen 31. UK Prospective Diabetes Study Group. Efﬁcacy of O. Effect of a multifactorial intervention on mortal- atenolol and captopril in reducing risk of macrovascular ity in type 2 diabetes. N Engl J Med 358: 580–91, and microvascular complications in type 2 diabetes 2008. UKPDS 39. Br Med J 317: 713–20, 1998. 20 Treatment of Dyslipidaemia

Peter Fasching 5th Medical Department, Wilhelminenspital, Vienna, Austria

patients are major challenges in preventive care Key messages: [6]. • Older patients with diabetes are often undertreated in terms of lipid-lowering, and may be at risk of preventable cardiovascular disease. 20.2 Risk assessment and treatment • Standard assessments of cardiovascular risk in targets older patients may be misleading and underestimate this risk; under these circumstances the Categories of risk based on lipoprotein levels in adults threshold for lipid lowering may need to be with diabetes mellitus according to most international adjusted. recommendations are given without modification con- • Statins are effective in significantly reducing cerning age and duration of diabetes [7, 8] (Table 20.1). cardiovascular risk in older patients aged up Since general cardiovascular risk is increasing with to 80 years with type 2 diabetes, but the effect both variables – and especially age – the cardiovascu- on mortality is inconsistent. lar risk in elderly diabetic and non-diabetic patients is generally underestimated according to non-age-specific risk assessment [9]. A more specific approach is cal- 20.1 Introduction culating individual risk by risk tables on the basis of epidemiological data [10]. Elderly and old patients with diabetes mellitus Principally, target values for treatment decisions suffer from increased cardiovascular risk [1–3]. based on LDL-cholesterol level in adults should be Coronary heart disease is a major cause of physical adopted without age limitation, especially in otherwise disability, particularly in the rapidly growing healthy and independent individuals (‘single disease population of elderly persons [4]. Cross-sectional model’) and yet, particular considerations for old (>75 data concerning lipid management, however, show years) and very old (>85 years) patients are advisable that patients with diabetes – and especially those in case of multimorbidity, dependency and end-stage of advanced age – are strikingly undertreated for dementia (‘multi-disease model’) due to limited life reasons apparently not related to LDL-cholesterol expectancy [11]. This may also involve the conse- levels [5]. Thus, special attention to cardiovas- quences of competing non-cardiovascular causes for cular risk factor management is obligatory, as death such as cancer and infections [12, 13]. In these the prevention of subsequent coronary events and cases, lipid-lowering therapy should be initiated only the maintenance of physical functioning in such on the basis of special individual causes.

Table 20.1 Treatment decisions based on LDL-cholesterol level in adult with diabetes according to the American Diabetes Association. • Lifestyle modiﬁcation focusing on the reduction of saturated fat, trans fat and cholesterol intake; weight loss; increased physical activity should be recommended to improve the lipid proﬁle in patients with diabetes. • Statin therapy should be added to life-style therapy, regardless of baseline lipid levels, for diabetic patients: • with overt cardiovascular disease (CVD); or • without CVD who are aged >40 years and have one or more other CVD risk factors. • For patients at lower risk than those mentioned above (e.g. without overt CVD and aged <40 years), statin therapy should be considered in addition to lifestyle therapy if LDL-cholesterol remains >100 mg dl−1, or in those with multiple CVD risk factors. • In individuals without overt CVD, the primary goal is an LDL-cholesterol <2.6 mmol l−1 (100 mg dl−1). • In individuals with overt CVD, a lower LDL-cholesterol goal of <1.8 mmol/l (70 mg dl−1),usingahighdoseofa statin, is an option. • If drug-treated patients do not reach the above targets on maximal tolerated statin therapy, a reduction in LDL-cholesterol of ∼40% from baseline is an alternative therapeutic goal. • Triglyceride levels <1.7 mmol l−1 (150 mg dl−1) and HDL-cholesterol levels >1.0 mmol l−1 (40 mg dl−1) in men and >1.3 mmol l−1 (50 mg dl−1) in women are desirable. However, LDL-cholesterol-targeted statin therapy remains the preferred strategy. • Combination therapy using statins and other lipid-lowering agents may be considered to achieve lipid targets, but this has not been evaluated in outcome studies for either CVD outcomes or safety.

20.3 Available evidence from men than in women [16] in the Cholesterol Treatment Trialists Collaborators (CTTC) (a systematic prospec- randomized controlled trials tive meta-analysis which reported data from 14 randomized trials) – those aged >65 years (n = 6446) had High cholesterol levels in older subjects are associated a 19% reduction in the risk of major cardiovascular with increased cardiovascular events and mortality, and events, a benefit similar to the 22% reduction in risk indeed, there is a positive relationship between the experienced by those aged <65 years (n = 7902) [17]. serum cholesterol level and cardiovascular risk. Statins Statins reduce the proportional risk as effectively are effective in lowering serum cholesterol levels, and in older as in younger people, although only lim- reducing the risk of cardiovascular events. There is no ited data are available for elderly patients with type evidence of a threshold below which lower cholesterol 2 diabetes. In the CTTC meta-analysis, which in- levels are not associated with a lower risk, which cluded 18 686 patients with diabetes among a total of suggests that cholesterol reduction is useful for all 90 056 participants, there was a 21% reduction (95% individuals at high cardiovascular risk, regardless of CI: 19–23%) in major vascular events per 1 mmol l−1 their baseline cholesterol level. reduction in serum LDL-cholesterol level, and no dif- The evidence for cholesterol lowering is evident for ference in treatment effect between patients with and individuals up to the age of 80 years. For those aged without diabetes [17]. >80 years, there is some evidence of benefit from Scientific evidence for the clinical effects of lipid observational studies (e.g. [14]). No mortality bene- lowering in elderly diabetic and non-diabetic patients fit was found for those aged >80 years who received on the basis of large randomized controlled trials a statin, whereas those aged 65–79 years showed a (RCTs) is therefore scarce. In the meantime, the de- significant 11% reduction in mortality. There was ev- tails of two statin trials – the HPS [18] and PROSPER idence, however, of a trend towards benefit in those [19] trials – have been reported which have included aged 80–85 years compared to those aged >85 years increased numbers of older patients with diabetes. [15]. Although the positive association between to- Statin therapy based on once-daily simvastatin or tal and LDL-cholesterol and cardiovascular risk be- pravastatin (40 mg) has been shown to significantly re- comes attenuated with advancing age – and more so in duce the relative risk (RR) of combined cardiovascular 20.3 AVAILABLE EVIDENCE FROM RANDOMIZED CONTROLLED TRIALS 283 endpoints and coronary events to a similar extent (ca. status post-percutaneous coronary intervention [23]; 20%; range 15–25%) in older (>70 years) and diabetic multivessel coronary disease [24]). subcohorts in both the HPS and PROSPER studies. In contrast to a reduction of all-cause-mortality by 12.9% 20.3.3 Fibrate therapy in the HPS study, the PROSPER trial found no ef- Because of the limited number of older diabetic sub- fect on total mortality. In the HPS study, however, the jects included in RCTs with gemfibrozil or fibrates, it therapeutic effects could be discriminated at the earliest is not possible to draw definite conclusions concern- after one year of treatment. ing hard clinical endpoints for this special group of In the CARDS trial, the efficacy of atorvastatin patients [25, 26]. (10 mg daily) versus placebo was examined in 2838 In the FIELD study, which included 9795 partic- type 2 diabetic patients aged 40–75 years. Unfortu- ipants aged 50–75 years with type 2 diabetes mel- nately, however, the study was terminated 2 years litus who were not receiving statin therapy at study earlier than expected after a median follow-up of 3.9 entry, micronized fenofibrate (200 mg daily) caused years, because the prespecified early stopping rule for no significant reduction in the risk of primary out- efficacy had been met. Here, a 10 mg daily dose of come of coronary events compared to placebo after atorvastatin reduced all major cardiovascular events by the 5-year study duration [27]. Yet, fenofibrate was = 37% (p 0.001), acute coronary heart disease events shown to reduce total cardiovascular events (–11%, by 36%, coronary revascularizations by 31%, the rate p = 0.05), due mainly to fewer non-fatal myocardial = of stroke by 49%, and mortality rate by 27% (p infarctions (−24%, p = 0.010) and revascularization 0.059) [20]. (−21%, p = 0.003). Mortality due to coronary heart A review of several large-scale clinical trials assess- disease was increased by 19%, albeit non-significantly. ing the efficacy of atorvastatin in the primary and sec- In conclusion, at present no firm recommendation can ondary prevention of cardiovascular events in patients be provided for the use of fibrate as first-line therapy with diabetes mellitus and/or metabolic syndrome, un- in diabetic patients. derlined the integrative place of atorvastatin in this Evidence-based clinical guidelines focusing on el- indication [21]. However, no differentiation with re- derly people [28] have presented recommendations for gards to patients of different ages was presented. lipid lowering in older patients with type 2 diabetes. The effect of statin therapy on the prevention of Their emphasis is, quite rightly, on good glycaemic strokes is heterogeneous, and consequently no general control, statin therapy for all those with abnormal lipid recommendation can yet be provided for the primary profiles, fibrate therapy for those with persistent raised prevention of stroke besides the secondary prevention triglycerides (≥2.3 mmol l−1), and referral to a special- of coronary disease. ist lipid or diabetes clinic for those with triglyceride levels >10 mmol l−1. 20.3.1 Statin therapy and adverse side effects 20.3.4 Other pharmacological therapies Statin therapy with simvastatin and pravastatin (40 mg, Until now, only one intervention study with hard once daily) is well tolerated in older and diabetic peo- clinical end points has been reported with the ple (partly on multiple drug therapy) (HPS, PROS- cholesterol-resorption inhibitor, ezetimibe. Since PER). In the CARDS study, no excess of adverse hyperlipdaemia is suggested as a risk factor for events was noted in the atorvastatin group when com- stenosis of the aortic valve, 1873 patients (mean pared to the placebo group [20]. age 67.5 years) with mild-to-moderate asymptomatic aortic stenosis were included in the SEAS trial, and 20.3.2 Statin therapy and special clinical randomized to daily treatment with either 40 mg simvastatin plus 10 mg ezetimibe, or placebo [29]. scenarios After a mean follow-up period of 52.2 months, the A number of smaller RCTs have suggested partic- primary endpoint (a composite of major cardiovascular ular beneficial effects of statin therapy in special events) was similar in both groups. Although there was clinical scenarios, which might have implications for no change in the need for aortic valve replacement, the treatment of older patients with diabetes mellitus fewer patients in the simvastatin-ezetimibe group had (e.g. unstable angina and non-Q-wave infarction [22]; ischaemic cardiovascular events (−22%, p = 0.02), 284 CH 20 TREATMENT OF DYSLIPIDAEMIA due mainly to the smaller number of patients who 2. Brun E, Nelson RG, Benett PH, Imperatore G, Zoppini underwent coronary artery bypass grafting. G, Verlato G and Muggeo M. Diabetes duration and A large-scale, randomized double-blind study in- cause-specific mortality in the Verona Diabetes Study. volving 15 067 patients at high cardiovascular risk Diabetes Care 23: 1119–23, 2000. was conducted to evaluate the effect of a combina- 3. Bertoni AG, Krop JS, Anderson GF and Brancati FL. tion of atorvastatin and the cholesteryl ester trans- Diabetes-related morbidity and mortality in a national fer protein (CETP) inhibitor torcetrapib, compared sample of U.S. elders. Diabetes Care 25: 471–5, 2002. to atorvastatin alone [30]. While, in the torcetrapib 4. Pinsky JL, Jette AM, Branch LG, Kannel WB and Fein- group, the HDL-cholesterol was increased by 72% and leib M. The Framingham Disability Study: relationship LDL-cholesterol decreased by 25% compared to base- of various coronary heart disease manifestations to dis- line, after a 12-month treatment period there was a ability in older persons living in the community. Am J Public Health 80: 1363–7, 1990. significantly increased risk of cardiovascular events 5. Massing MW, Sueta CA, Chowdhury M, Biggs DP and (+25%, p = 0.001) and death from any cause (+58%, Simpson RJ. Lipid management among coronary artery p = 0.006), albeit of unknown mechanism. disease patients with diabetes mellitus or advanced age. Am J Cardiol 87: 646–9, 2001. 20.4 Eligibility for lipid-lowering 6. Ades PA. Cardiac rehabilitation and secondary preven- therapy tion of coronary disease. N Engl J Med 345: 892–902, 2001. 7. American Diabetes Association ADA: Standards of Lipid lowering in older patients with diabetes must Medical Care - 2008. Dyslipdemia/lipid management: be seen as one part of the multifactorial intervention Recommendations. Diabetes Care 31 (Suppl. 1): S26–7, process necessary in diabetes management to prevent 2008. and/or delay cardiovascular complications [31, 32]. 8. Executive Summary of the Third Report of the National Lipid-lowering therapy in patients with type 2 dia- Cholesterol Education Program NCEP Expert Panel on betes seems the case for early intervention. Yet, in view Detection, Evaluation and Treatment of High Blood of the divergent study results and outstanding data, an Cholesterol in Adults (Adult Treatment Panel III). assessment of the risk of the individual with type 2 di- JAMA 285: 2486–97, 2001. abetes is mandatory to assist clinical decision-making 9. D’Agostino RB, Russell MW, Huse DM, et al. (2000) when initiating lipid therapy [33]. Primary and subsequent coronary risk appraisal∼:new The denial of an otherwise recommended results from the Framingham study. Am Heart J139: 272. lipid-lowering therapy in older patients with diabetes 10. Rutten GEHM, Verhoeven S, Heine RJ, De Grauw mellitus is not justified on the basis of present WJC, Cromme PVM, Reenders K, Van Ballegooie, knowledge, if the patient is principally eligible for Wiersma TJ and the St Vincent Declaration Primary invasive procedures such as coronary angiography, Care Diabetes Group. (1999) Approved Guidelines for percutaneous transluminal coronary angiography and Type 2 Diabetes 2002. Huisarts en Wetenschap 42 (2): stenting for coronary heart disease or surgery (i.e. 67–84. coronary bypass or heart valve surgery, or carotid 11. Sinclair AJ. Diabetes in old age: changing concepts in endarterectomy). the secondary care arena. J Roy Coll Phys 2000; 34: Finally, in very frail individuals, great caution must 240–4. be exercised before embarking on lipid lowering with 12. Weverling-Rijnsburger AW, Blauw GJ, Lagaay AM, a statin, largely because of the likely increased risk of Knook DL, Meinders AE and Westendorp RG. Total muscle disease-related side effects. Patient safety and cholesterol and risk of mortality in the oldest old. balancing the risks and benefits are important priorities Lancet 350: 119–23, 1997. when managing older patients, and are hallmarks of the 13. Schatz IJ, Masaki K, Yano K, Chen R, Rodriguez BL and Curb JD. Cholesterol and all-cause mortality in careful clinician. elderly people from the Honolulu Heart Program: a cohort study. Lancet 358: 351–5, 2001. References 14. Aronow WS and Ahn C. Incidence of new coronary events in older persons with prior myocardial infarction 1. Sinclair AJ, Robert IE and Croxon SCM. Mortality in and serum low-density lipoprotein cholesterol ≥125 older people with diabetes mellitus. Diabetic Med 14: mg/dl treated with statins versus no lipid-lowering 639–47, 1997. drug. Am J Cardiol. 2002; 89: 67–9. 20.4 ELIGIBILITY FOR LIPID-LOWERING THERAPY 285

15. Foody JM, Rathore SS, Galusha D, et al.. Hydroxy- 24. Pitt B, Waters D, Brown WV, et al. for the Atorvastatin methylglutaryl-CoA reductase inhibitors in older per- versus Revascularization Treatment Investigators. Ag- sons with acute myocardial infarction: evidence for an gressive lipid-lowering therapy compared with angio- age statin interaction. J Am Geriatr Soc. 2006; 54: plasty in stable coronary artery disease. N Engl J Med 421–30. 341: 70–6, 1999. 16. Anum EA and Adera T. Hypercholesterolemia and 25. Rubins HB, Robins SJ, Collins D, Fye CL, Andersen coronary heart disease in the elderly: a meta-analysis. JW, Elam MB, Faas FH, Linares E, Schaefer EJ, Ann Epidemiol. 2004; 14: 705–21. Schectman G, Wilt TJ, Wittes J. Gemfibrozil for the 17. Baigent C, Keech A, Kearney PM, et al. The Choles- secondary prevention of coronary heart disease in men terol Treatment Trialists’ (CTT) Collaborators: Efficacy with low levels of high-density lipoprotein cholesterol: and safety of cholesterol-lowering treatment: prospec- Veterans Affairs High-Density Lipoprotein Cholesterol tive meta-analysis of data from 90,056 participants Intervention Trial Study Group. N Engl J Med 341: in 14 randomised trials of statins. Lancet 2005; 366: 410–18, 1999. 1267–78. 26. Diabetes Atherosclerosis Intervention Study In- 18. Heart Protection Study Collaborative Group. vestigators. Effect of fenofibrate on progression of MRC/BHF Heart Protection Study of cholesterol coronary-artery disease in type 2 diabetes: the Diabetes lowering with simvastatin in 20 536 high-risk Atherosclerosis Intervention Study, a randomised individuals: a randomised placebo-controlled trial. study. Lancet 357: 905–10, 2001. Lancet, 360, 7–22, 2002. 27. Keech A, Simes RJ, Barter P, Best J, Scott R, Taskinen 19. Shepherd J, Blauw GJ, Murphy MB, et al. On behalf MR, Forder P, Pillai A, Davis T, Glasziou P, Drury of the PROSPER study group. Pravastatin in elderly P, Kesaniemi¨ YA, Sullivan D, Hunt D, Colman P, individuals at risk of vascular disease (PROSPER): d’Emden M, Whiting M, Ehnholm C, Laakso M and the a randomised controlled trial. Lancet 360: 1623–30, FIELD study investigators. Effects of long-term fenofi- 2002. brate therapy on cardiovascular events in 9795 people 20. Colhoun HM, Betteridge DJ, Durrington PN, Hitman with type 2 diabetes mellitus (the FIELD study): ran- GA, Neil HA, Livingstone SJ, Thomason MJ, Mack- domised controlled trial. Lancet 366: 1849–61, 2005. ness MI, Charlton-Menys V, Fuller JH and the CARDS 28. The European Diabetes Working Party for Older Peo- investigators. Primary prevention of cardiovascular ple, 2001-2004. Clinical Guidelines for Type 2 Dia- disease with atorvastatin in type 2 diabetes in the betes Mellitus. (available on www.eugms.org). Collaborative Atorvastatin Diabetes Study (CARDS): 29. Rossebo AB, Pedersen TR, Boman K, et al.forthe multicentre randomised placebo-controlled trial. Lancet SEAS Investigators. Intensive lipid lowering with sim- 364: 685–96, 2004. vastatin and ezetimibe in aortic stenosis. N Engl J Med 21. Arca M. Atorvastatin efficacy in the prevention of 359: 1343–56, 2008. cardiovascular events in patients with diabetes melli- 30. Barter PJ, Caulfield M, Eriksson M, et al.forthe tus and/or metabolic syndrome. Drugs 67 (Suppl. 1): ILLUMINATE Investigators. Effects of torcetrapib in 43–54, 2007. patients at high risk for coronary events. N Engl J Med 22. Waters DD, Schwartz GG, Olsson AG, et al.forthe 357: 2109–22, 2007. MIRACLE Study Investigators. Effects of atorvas- 31. Gaede P, Vedel P, Larsen N, Jensen GVH, Parving tatin on stroke in patients with unstable angina or HH and Pedersen O. Multifactorial intervention and non-Q-wave myocardial infarction. A myocardial is- cardiovascular disease in patients with type 2 diabetes. chemia reduction with aggressive cholesterol lowering N Engl J Med 348: 383–93, 2003. (MIRACLE) substudy. Circulation 106: 1690–5, 2002. 32. Gaede P, Lund-Andersen H, Parving HH and Pedersen 23. Serruys PWC, de Feyter P, Macaya C, et al.forthe O. Effects of a multifactorial intervention on mortality Lescol Intervention Prevention Study (LIPS) Inves- in type 2 diabetes. N Engl J Med 358: 580–91, 2008. tigators. Fluvastatin for prevention of cardiac events 33. Steinmetz A. Lipid-lowering therapy in patients with following successful first percutaneous coronary inter- type 2 diabetes: the case for early intervention. Diabetes vention. A randomised controlled trial. JAMA 287: Metab Res Rev 24: 286–93, 2008. 3215–22, 2002. 21 Hypoglycaemia

Vincent McAulay1 and Brian M. Frier2 1Department of Diabetes, Crosshouse Hospital, Kilmarnock, UK 2Department of Diabetes, Royal Inﬁrmary, Edinburgh, UK

major vascular event such as myocardial infarction and Key Messages stroke as a consequence of hypoglycaemia. In addition, frequent and unpredictable hypoglycaemia in an old • The frequency and severity of hypoglycaemia person, causing problems such as dizziness, disturbed in the elderly is commonly underestimated, and balance, weakness, transient loss of consciousness and often mis-attributed to other conditions. falls, can undermine their self-confidence and have a • The symptoms of hypoglycaemia are different in destabilizing effect on their independence and ability the elderly compared to younger people, partly to live alone. The imposition of the burden of pre- as a function of ageing. vention and treatment of hypoglycaemia on relatives • The increasingly stringent glycaemic targets inherent in modern diabetes care demands that and carers, many of whom may also be elderly, may patients, and their relatives and carers, are fully provoke further domestic difficulties, threatening the informed about the symptoms and effects of independence of the affected individual, and possi- hypoglycaemia and its emergency management. bly precipitating their transfer to residential care. The emergence of hypoglycaemia as a prominent problem in an elderly person with diabetes may influence their management by encouraging health attendants to avoid 21.1 Introduction the risk of exposure to a low blood glucose and to accept poor or suboptimal glycaemic control. Hypoglycaemia is the most serious and disruptive side Most people who develop diabetes later in life have effect of the treatment of diabetes in the elderly. While type 2 diabetes, the prevalence of which is rising in it is a recognized clinical consequence of the use of association with increasing longevity, and treatment insulin and sulphonylureas in all age groups requir- with insulin is increasing in elderly people as a conse- ing such therapy, the frequency of hypoglycaemia is quence of the decline in pancreatic beta-cell function underestimated in elderly people with diabetes. This that occurs as diabetes progresses in severity. In addi- may be because its clinical manifestations are unrec- tion, some older people have an insulin-deficient form ognized or are wrongly attributed to other pathological of diabetes, and require insulin from the time of diag- conditions such as cerebral ischaemia or degenera- nosis, while many more people with type 1 diabetes tive disorders. The effects of hypoglycaemia can be now survive into old age. Although hypoglycaemia severe in elderly patients, many of whom are physi- was considered to be a relatively uncommon side ef- cally frail and have coexisting macrovascular disease. fect of oral antidiabetic agents [1], many sulphony- They may therefore be at increased risk of suffering a lureas promote hypoglycaemia of varying frequency

and severity. The increasing popularity of combining ↓ insulin oral antidiabetic agents with insulin in the treatment of ∼4.6 mmol 1−1 type 2 diabetes may possibly augment the magnitude of this problem, particularly in the elderly patient. The glucose glucose risk of exposure to insulin-induced hypoglycaemia is, ↑counter-regulation therefore, rising steadily in older people with diabetes. • glucagon • epinephrine Before examining the epidemiology and morbid- ∼3.8 mmol 1−1 ity of hypoglycaemia in the elderly population with

diabetes, it is important to ascertain whether age autonomic per se affects the symptomatic awareness of – and symptoms −1 counter-regulatory hormonal response to – acute hy- ∼3.0 mmol 1 poglycaemia, and how these fundamental responses to a low blood glucose may be modiﬁed by the presence cognitive of type 2 diabetes in an ageing population. dysfunction ∼2.8 mmol 1−1

21.2 Physiological responses to Figure 21.1 Hierarchy of responses to hypoglycaemia hypoglycaemia in non-diabetic humans.

Hypoglycaemia 21.2.1 Counter-regulation The human brain is dependent upon glucose as its prin- HYPOTHALAMUS cipal source of fuel, and requires a continuous supply PITUITARY Growth hormone of glucose via the cerebral circulation. Depriving the brain of glucose rapidly causes neuroglycopenia which has various effects, including impairment of cognitive ACTH function. LIVER

In humans, several mechanisms have evolved to Autonomic maintain glucose homeostasis and so protect the in- nervous system tegrity and function of the brain [2]. A decline in blood Hypoglycaemia glucose concentration activates a characteristic hierar- ADRENAL chy of responses, commencing with the suppression of endogenous insulin secretion, the release of several PANCREAS counter-regulatory hormones, and the subsequent de- Norepinephrine velopment of characteristic symptoms (Figure 21.1). Epinephrine Cortisol Glucagon These alert the informed individual to the development of hypoglycaemia, so allowing them to take Figure 21.2 Principal components of glucose counter- early and appropriate action (the ingestion of carbo- regulation in humans. hydrate) to assist recovery. Such protective responses are usually effective in maintaining the arterial blood also by the generation of characteristic autonomic glucose concentration within a normoglycaemic range warning symptoms (Figure 21.2). Although glucagon (which can be arbitrarily defined as a blood glu- is the most potent counter-regulatory hormone, the cose above 3.8 mmol l−1), which protects the brain role of epinephrine (adrenaline) becomes paramount from exposure to neuroglycopenia caused by pro- if the secretory response of glucagon is deficient [3]. tracted glucose deprivation. Glucose counter-regulation Other counter-regulatory hormones, such as cortisol is controlled from centres within the brain (mainly and growth hormone, have greater importance in pro- the ventromedial hypothalamus), assisted by activa- moting recovery from prolonged hypoglycaemia. tion of hypothalamic autonomic nervous centres with Both, glucagon and epinephrine stimulate hepatic stimulation of the peripheral sympathoadrenal sys- glycogenolysis, releasing glucose from glycogen stored tem. This contributes to glucose counter-regulation in the liver, and also promote gluconeogenesis from through the peripheral actions of catecholamines, and three-carbon precursors such as alanine, lactate and 21.2 PHYSIOLOGICAL RESPONSES TO HYPOGLYCAEMIA 289 glycerol. The energy for this process is provided by the CEREBRAL CORTEX hepatic oxidation of free fatty acids that are released by lipolysis. Catecholamines inhibit insulin secretion, Neuroglycopenic diminish the peripheral uptake of glucose, stimulate symptoms lipolysis and proteolysis and promote glycogenolysis in peripheral muscle to provide lactate, which is utilized Hypoglycaemia for gluconeogenesis in the liver and kidney. Hypothalamus 21.2.2 Symptoms Both the sympathetic and parasympathetic divisions of the autonomic nervous system are activated during hy- Parasympathetic poglycaemia, leading to the direct neural stimulation nervous system Sympathetic of end-organs via peripheral autonomic nerves, and the nerves physiological effects are augmented by the secretion of epinephrine from the adrenal medulla [4] (Figure 21.3). Adrenal Studies in young adults using physiological and phar- medulla macological methods to assess the symptoms of hypoglycaemia, have confirmed that the symptoms of Adrenaline pounding heart, tremulousness and feeling nervous or Tremor Heart Sudomotor anxious are adrenergic in nature [5]. The sweating re- eccrine sponse to hypoglycaemia is mediated primarily via sweat glands sympathetic cholinergic stimulation [5, 6], with cir- Pounding heart culating catecholamines possibly contributing through Autonomic Sweating the activation of α-adrenoceptors [7]. When the brain symptoms is deprived of glucose, it rapidly malfunctions, caus- Perception, ing interference with information processing and the interpretation development of cognitive dysfunction, which under- and action lies the generation of neuroglycopenic symptoms such Figure 21.3 Activation of the autonomic nervous as difficulty concentrating, feelings of tiredness and system and the sympathoadrenal system during hypo- drowsiness, faintness, dizziness, generalised weakness, glycaemia. Reproduced with permission from Frier, B.M. confusion, difficulty speaking and blurring of vision. and Fisher, B.M. (eds) (1993), Hypoglycaemia and Dia- Statistical techniques have also been used to clas- betes, Edward Arnold (Publisher) Ltd, London. sify the symptoms of hypoglycaemia. Applying methods such as Principal Components Analysis (PCA), Table 21.1 Classification of symptoms of hypo- the symptoms of hypoglycaemia segregate into three glycaemia in patients with insulin-treated diabetes distinct factors or groups: neuroglycopenic, autonomic depending on age group. and general malaise [8]. This ‘three-factor’ validated model containing 11 common symptoms of hypogly- Children Adults Elderly caemia (the ‘Edinburgh Hypoglycaemia Scale’), has (pre-pubertal) been used to classify symptoms objectively in various Autonomic/ Autonomic Autonomic groups of subjects, and has shown age-specific dif- Neuroglycopenic ferences in the nature of hypoglycaemic symptoms as Neuroglycopenic Neuroglycopenic classified by this statistical method (Table 21.1). Symp- Behavioural Non-specific Neurological toms of hypoglycaemia are idiosyncratic and vary be- malaise tween individuals. They may also differ in intensity Reproduced with permission from Ref. [10]. in different situations, and their perception can be influenced by distraction or other external influences. In rather than the number or nature of the symptoms perceiving the onset of hypoglycaemia (often described generated. An assessment of the subjective reality of as subjective ‘awareness’), the intensity of a few car- symptoms is therefore essential in attempting any form dinal symptoms is of importance to the individual, of measurement or devising a scoring system. Both 290 CH 21 HYPOGLYCAEMIA autonomic and neuroglycopenic symptoms appear to be of equal value in warning people with type 1 Counter-regulatory diabetes of the onset of hypoglycaemia, provided that symptons the symptoms peculiar to the individual are identified and interpreted correctly [11]. hormones

21.2.3 Glycaemic thresholds Speciﬁc physiological responses occur when the de-

clining blood glucose reaches different levels of hy- Magnitude of response poglycaemia. Although these glycaemic thresholds are readily reproducible in non-diabetic humans [12], they are plastic and dynamic and can be modified. In non-diabetic humans the glycaemic threshold at which Hypo 1 Hypo 2 the secretion of most counter-regulatory hormones is Time (hours - days) triggered is around 3.8 mmol l−1 (measured as arterialized blood glucose), so that counter-regulation Figure 21.5 Schematic representation of the effect of is usually activated when blood glucose falls below antecedent hypoglycaemia on the neuroendocrine and the normal range. Counter-regulation therefore oc- symptomatic responses to subsequent hypoglycaemia. curs at a blood glucose that is higher than that at Reproduced with permission from BM Frier and M Fisher (eds) (2007), Hypoglycaemia in Clinical Practice, John which the symptomatic response to hypoglycaemia oc- Wiley & Sons Ltd, Chichester. curs (3.0 mmol l−1) and before the onset of cognitive dysfunction (2.8 mmol l−1) (Figure 21.1). The glycaemic threshold for autonomic symptoms coincides hormonal and symptomatic responses to hypogly- with the classical autonomic ‘reaction’ to hypoglycaemia do not occur until much lower blood glu- caemia, which can be identified by the sudden devel- cose levels are reached, particularly when the glycated opment of physiological changes [13]. haemoglobin concentration is within the non-diabetic In people with diabetes, glycaemic thresholds can be range [14]. Similarly, antecedent hypoglycaemia last- modified by the prevailing glycaemic state, and partic- ing for 1 h or more has been shown to diminish the ularly by strict control (Figure 21.4), and can be influenced by metabolic perturbations such as preceding magnitude of the symptomatic and neuroendocrine (antecedent) hypoglycaemia. Many studies in people responses to any subsequent episode of hypogly- with insulin-treated diabetes who have strict glycaemic caemia occurring within the following 24–48 h [13] control have demonstrated that the counter-regulatory (Figure 21.5). This duration of untreated hypoglycaemia is not uncommon during sleep, and may be one

STRICT GLYCAEMIC CONTROL of the mechanisms that induce impaired awareness of hypoglycaemia in people with type 1 diabetes. )

1 COUNTER-REGULATION

− 3.8

3.0 SYMPTOMS 21.3 Acquired hypoglycaemia syndromes in type 1 diabetes Arterialized blood glucose (mmol 1 21.3.1 Counter-regulatory deficiencies In many people with type 1 diabetes, the glucagon Figure 21.4 The glycaemic thresholds for counter- regulatory hormonal secretion and the onset of symptoms secretory response to hypoglycaemia becomes dimin- can vary depending on the prevailing level of glycaemic ished or absent within a few years of the onset of control in people with diabetes. Strict glycaemic control insulin-deficient diabetes. With glucagon deficiency is associated with a higher glycaemic threshold (i.e. alone, blood glucose recovery from hypoglycaemia a lower blood glucose concentration is required), is relatively unaffected because counter-regulation is providing a more intense hypoglycaemic stimulus. maintained by the actions of epinephrine. However, 21.3 ACQUIRED HYPOGLYCAEMIA SYNDROMES IN TYPE 1 DIABETES 291

Table 21.2 Frequency of abnormal counter-regulatory responses to hypoglycaemia in patients with type 1 diabetes. Duration of Glucagon Adrenaline Cortisol Growth hormone diabetes (years) (%) (%) (%) (%) <127900 1–5 75 25 0 0 5–10 100 44 11 11 >10 92 66 25 25

Reproduced with permission from Frier, B.M. and Fisher, B.M. (eds) (1999), Hypoglycaemia and Diabetes, Edward Arnold (Publisher) Ltd. Table 21.3 Factors influencing normal awareness of in up to 45% of people who have type 1 dia- hypoglycaemia. betes of long duration, a dual impairment of the secretion of glucagon and epinephrine is observed Internal External [15], predisposing them to serious deficiencies of glu- Physiological Drugs cose counter-regulation when exposed to hypogly- Recent glycaemic control Beta-adrenoceptor blockers caemia, delaying the recovery of blood glucose, and (non-selective) allowing progression to more severe hypoglycaemia Degree of neuroglycopenia Hypnotics, tranquilizers (Table 21.2). People with type 1 diabetes of long dura- Symptom Alcohol tion are therefore at increased risk of developing severe intensity/sensitivity and prolonged hypoglycaemia, particularly when in- Psychological Environmental tensive insulin therapy is used [16]. Indeed, people Arousal Sleep with type 1 diabetes who have combined deficiencies Focused attention Posture Congruence; denial Distraction of their glucagon and epinephrine responses to hypo- Competing explanations glycaemia have been shown to be at 25-fold or even higher increased risk for severe iatrogenic hypogly- Education caemia if subjected to intensive insulin therapy com- Knowledge Symptom belief pared to those whose glucagon response is absent but who retain their epinephrine response [16, 17]. These Reproduced with permission from Ref. [13]. counter-regulatory deficiencies co-segregate with impaired awareness of hypoglycaemia in people with type Impaired awareness of hypoglycaemia is generally 1 diabetes [18], suggesting that they share a common thought to result from diminished sympathoadrenal ac- pathogenetic mechanism within the brain. tivation with a resultant reduction in the symptomatic response to a given level of hypoglycaemia [19]. This acquired syndrome of hypoglycaemia is common in 21.3.2 Impaired awareness of type 1 diabetes, affecting around one-quarter of pa- hypoglycaemia tients, becomes more prevalent with increasing dura- This occurs when the symptomatic warning is dimin- tion of diabetes, and predisposes the patient to a much ished or inadequate in people with diabetes, and many higher risk of severe hypoglycaemia than people who factors can influence the awareness of hypoglycaemia retain normal awareness [13]. It is much less com- (Table 21.3). Impaired awareness is not an ‘all-or-none’ mon in insulin-treated patients with type 2 diabetes phenomenon. Rather, a ‘partial’ impairment of aware- [20]. In some patients with type 1 diabetes, when im- ness may develop, with the individual being aware paired awareness of hypoglycaemia is associated with of some episodes of hypoglycaemia, but not others. strict glycaemic control during intensive insulin ther- Alternatively, they may experience a reduction in the apy, or has followed episodes of recurrent severe hy- intensity or number of symptoms which varies between poglycaemia [21], this syndrome may be reversible hypoglycaemic events, and progress to ‘absent’ aware- by relaxing glycaemic control or by avoiding further ness where the patient is no longer aware of the onset hypoglycaemia, but in many patients with type 1 di- of hypoglycaemia. Several mechanisms underlying this abetes of long duration, it appears to be a permanent problem have been proposed (Table 21.4). defect [13]. 292 CH 21 HYPOGLYCAEMIA

Table 21.4 Possible mechanisms of impaired aware- study of the counter-regulatory hormonal responses to ness of hypoglycaemia. hypoglycaemia induced by an intravenous infusion of insulin suggested that diminished secretion of growth CNS adaptation hormone and cortisol is a feature of advanced age, and Chronic exposure to low blood glucose a modest impairment of hormonal counter-regulatory Strict glycaemic control in diabetic patients secretion was associated with some attenuation of Insulinoma in non-diabetic patients blood glucose recovery [25]. Insulin clearance was re- Recurrent transient exposure to low blood glucose duced, as was the secretion of glucagon, while the Antecedent hypoglycaemia release of epinephrine was delayed; these abnormalities were unchanged following a period of physical CNS glucoregulatory failure training, suggesting that they were not secondary to Counter-regulatory deficiency (hypothalamic defect?) Hypoglycaemia-associated central autonomic failure a sedentary lifestyle [25]. However, a different study using the hyperinsulinaemic glucose-clamp technique Peripheral nervous system dysfunction has suggested that age per se has no effect [26]. Com- peripheral autonomic neuropathy parative analysis and interpretation of these studies reduced peripheral adrenoceptor sensitivity are problematical because of differences between study groups with respect to the rate of onset and duration Reproduced with permission from Ref. [13]. of hypoglycaemia and of the magnitude of the plasma insulin concentrations achieved – factors which can in- 21.3.3 Central autonomic failure fluence the nature of the counter-regulatory hormonal Because hormonal counter-regulatory deficiencies response. and impaired awareness of hypoglycaemia usually A study in older non-diabetic subjects (mean age coexist, and are associated with an increased 76 years) [27], using a stepped glucose-clamp tech- frequency of severe hypoglycaemia, the concept of nique, demonstrated deficiencies in the secretion of a “Hypoglycaemia-Associated Autonomic Failure” glucagon and epinephrine. In another study [28], the (HAAF) was proposed [22], when it was argued that counter-regulatory responses in 11 older, non-diabetic recurrent severe hypoglycaemia is the primary problem individuals (mean age 65 years) were compared to which provokes these acquired abnormal responses. those of 13 young, healthy volunteers (mean age 24 The concept of HAAF deems that recent antecedent years). Subtle differences were observed in the magni- hypoglycaemia, in people with type 1 diabetes tude of the hormonal counter-regulatory responses in [22] and in patients with type 2 diabetes who have the older group (in whom the epinephrine, glucagon, progressed to pancreatic beta-cell failure [23], causes pancreatic polypeptide and cortisol responses were defective glucose counter-regulation in the setting of lower) in response to modest hypoglycaemia (arte- − an absent glucagon response. This occurs because rialized blood glucose 3.3 mmol l 1). However, no the epinephrine response is then markedly attenuated such differences were demonstrated when the hypo- during exposure to subsequent hypoglycaemia, while glycaemic stimulus was more profound (arterialized − impaired awareness of hypoglycaemia develops blood glucose 2.8 mmol l 1). Two further studies in through blunting of the sympathoadrenal response and non-diabetic elderly subjects, using similar designs and reduced generation of autonomic symptoms [13, 24]. methodologies, failed to demonstrate any significant age-related impairment of the counter-regulatory hormonal responses to hypoglycaemia [29, 30]. 21.4 Effects of age on physiological responses to hypoglycaemia 21.4.2 Symptomatic response to hypoglycaemia 21.4.1 Counter-regulatory mechanisms Differences between age groups in the symptom pro- Because, in humans, many physiological processes al- files in response to hypoglycaemia have been demon- ter with advancing age, it is important to determine strated in children and adults with type 1 diabetes whether the ageing process per se may affect the nature [9], while older people with insulin-treated type 2 and efficacy of the glucose counter-regulatory response diabetes have been observed to experience a clus- to hypoglycaemia. In non-diabetic elderly subjects, a ter of ‘neurological’ symptoms (unsteadiness, poor 21.4 EFFECTS OF AGE ON PHYSIOLOGICAL RESPONSES TO HYPOGLYCAEMIA 293

Table 21.5 Symptoms of hypoglycaemia in the elderly Table 21.7 Hypoglycaemia in the elderly: symptoms. [31]. 1. Autonomic symptoms are not selectively diminished Neuroglycopenic Autonomic Neurological 2. Intensity of all symptoms (historical reports and exper- Weakness Sweating Unsteadiness imental studies) is low Drowsiness Shaking Poor coordination 3. Glycaemic threshold for onset of symptoms is altered by Poor concentration Pounding heart Double vision Dizziness Anxiety Blurred vision age; lower blood glucose is required to initiate symptoms Confusion Slurred speech 4. Cognitive dysfunction induced simultaneously by hypo- Light-headedness glycaemia may interfere with perception of symptoms 5. Awareness of hypoglycaemia may be reduced by ageing Table 21.6 Hypoglycaemia in the elderly: effects of age. for the generation of symptoms is modiﬁed by age, 1. Mild attenuation of blood glucose recovery may occur with a lower blood glucose being required to initiate a (hepatic glucose production is diminished) symptomatic response [30] (Table 21.7). 2. Modest reductions demonstrable in counter-regulatory hormonal responses (but maximal response to more 21.4.3 Cognitive function severe hypoglycaemia) The hierarchy of the cognitive changes in response to 3. Symptom response is less intense with altered glycaemic hypoglycaemia may change with age. In one study of threshold and reduced awareness of hypoglycaemia non-diabetic subjects [30], the responses to moderate hypoglycaemia of seven elderly men were compared with those of seven young men. The four-choice re- coordination, slurring of speech and visual distur- action time, a measure of psychomotor coordination, bances) [31], in addition to the classical autonomic deteriorated in the older men at a mean (±SD) plasma and neuroglycopenic groups of symptoms recognized − glucose of 3.0 ± 0.1 mmol l 1, compared to 2.6 ± 0.1 in young adults (Table 21.5). Age per se may there- − mmol l 1 in the young group, and the abnormality fore modify the nature and intensity of some symptoms was more profound (Figure 21.6). The symptomatic re- of hypoglycaemia, possibly as a consequence of other sponse to hypoglycaemia commenced at a lower blood age-related changes such as effects on cerebral circu- glucose concentration in the older men than in the lation, and the presence of underlying cerebrovascular − young adults (3.0 ± 0.2versus3.6 ± 0.1 mmol l 1), disease or degenerative abnormalities of the central while in the older subjects the glycaemic threshold for nervous system (Table 21.6). subjective symptomatic awareness of hypoglycaemia In a small group of non-diabetic subjects in and that for the onset of cognitive dysfunction were whom hypoglycaemia was induced using a stepped glucose-clamp, lower symptom scores were recorded in the seven older subjects (mean age 72 years) than in 4.0 the six younger subjects (mean age 30 years), and the SYMPTOMS

) younger men 1 3.5 usual haemodynamic responses to hypoglycaemia (par- − ticularly a rise in heart rate) were absent in the older group [29]. This suggests that the symptomatic aware- 3.0 older men ness of hypoglycaemia may be reduced in the elderly, and is associated with an attenuated end-organ response younger men lucose (mmol 1

Arterialized blood 2.5 to sympathoadrenal stimulation. As these responses g REACTION TIME generate many of the autonomic symptoms of hypogly- (FOUR CHOICE) caemia, the perception of hypoglycaemia is affected. 2.0 In another study of older, non-diabetic subjects, the symptomatic response to hypoglycaemia commenced Figure 21.6 The difference between the glycaemic at a lower blood glucose (mean ± SD: 3.0 ± 0.2 threshold for subjective awareness of hypoglycaemia mmol l−1) compared to a younger group (3.6 ± 0.1 and that for the onset of cognitive dysfunction may be mmol l−1), suggesting that the glycaemic threshold absent in the elderly. Derived from data in Ref. [30]. 294 CH 21 HYPOGLYCAEMIA coincidental. A similar difference has been observed disparate methods used to induce hypoglycaemia [17, in patients with type 1 diabetes who have an impaired 35–37]. In these early studies the counter-regulatory awareness of hypoglycaemia, in whom the onset of the hormonal responses were either normal or only mildly cognitive dysfunction induced by hypoglycaemia either impaired in people with type 2 diabetes. Epinephrine preceded or was coincidental with the onset of a symp- secretion was invariably preserved (Table 21.8). tomatic response [13]. This observation suggests that More recent studies of counter-regulatory responses elderly people may be at an intrinsically greater risk of to hypoglycaemia in people with type 2 diabetes, developing neuroglycopenia because the onset of warn- treated either with diet or oral medication, have ing symptoms and cognitive impairment occur simul- shown that counter-regulatory hormone release occurs taneously, so interfering with their ability to recognize at higher blood glucose levels than in non-diabetic and take action to self-treat a fall in blood glucose. control subjects. In one study [41], hypoglycaemia was induced using a stepped glucose-clamp in 11 subjects with type 2 diabetes (mean age 56 years) 21.5 Effects of type 2 diabetes on and in a group of eight age- and weight-matched, responses to hypoglycaemia non-diabetic subjects. Other comparator groups comprised 10 subjects with type 1 diabetes and 21.5.1 Counter-regulation another non-diabetic control group. The secretion of counter-regulatory hormones occurred at a higher Good glycaemic control in type 2 diabetes limits the blood glucose level in the subjects with type 2 diabetes development and severity of vascular complications, than in those with type 1 diabetes, irrespective of but achieving this with insulin and many of the oral the quality of glycaemic control as measured by antidiabetic agents inevitably increases the risk of hy- glycated haemoglobin (HbA1). However, differences poglycaemia [1]. The counter-regulatory and symp- in the gender distribution and insulin infusion rates tomatic responses to hypoglycaemia have been studied in some subjects may have introduced potential in patients with type 2 diabetes, but earlier studies were confounders that could have influenced the results performed using a variety of techniques and proto- [41]. Non-diabetic female subjects with type 1 cols which makes comparisons between studies either diabetes have a lesser magnitude of counter-regulatory difficult or impossible. Problems included the study responses to hypoglycaemia than male subjects of heterogeneous groups of subjects with type 1 and [42–44], so an over-representation of male subjects in type 2 diabetes [32], variations in the magnitude of the first study [41] might have influenced the results. the hypoglycaemic stimulus between subjects because Similarly, in non-diabetic and type 1 diabetic subjects, the blood glucose differed at baseline [33, 34], and hyperinsulinaemia suppresses glucagon release and

Table 21.8 Studies of hormonal counter-regulation to hypoglycaemia in type 2 diabetes. Study No. of Method of hypogly- Mean glucose Hormonal response patients caemia induction nadir (mmol l−1) [33] 27 iv insulin bolus 2.0 Reduced glucagon [35] 10 iv insulin bolus 1.8 No impairment [36] 10 iv insulin infusion 1.7 No impairment [34] 8 iv insulin infusion 1.9 No impairment [17] 13 sc insulin bolus 3.4 Reduced glucagon, cortisol and growth hormone [37] 10 iv insulin infusion 2.4 No impairment [38] 10 iv insulin infusion 2.8 Reduced glucagon and growth hormone; increased adrenaline and cortisol [39] 9 iv insulin infusion 3.4 Reduced glucagon and increased epinephrine (adrenaline) [40] 7 iv insulin infusion 2.4 Glucagon response preserved in 5 patients; magnitude of adrenaline response increased if poor glycaemic control [41] 11 iv insulin infusion 2.2 No impairment 21.5 EFFECTS OF TYPE 2 DIABETES ON RESPONSES TO HYPOGLYCAEMIA 295 increases the secretion of catecholamines and cortisol 120 ) 1 in response to hypoglycaemia [45–48], such that the − 100 higher rate of insulin infusion required in some of 80 Non-Diabetic the subjects with type 2 diabetes might also have 60 Type 2 OHA Type 2 Insulin modified their results. Glucagon 40 Although type 2 diabetes may confer some protec- 20 concentration (pg ml tion against hypoglycaemia, especially in those sub- 0 jects with relatively poor glycaemic control, improving 0 −30 60 120 180 240 300 blood glucose control with insulin alters the threshold Time (min) for the counter-regulatory response to hypoglycaemia [40, 41], in a manner similar to that observed in people Figure 21.7 Mean plasma glucagon concentrations with type 1 diabetes when their glycaemic control is during hyperinsulinaemic stepped hypoglycaemic improved [19]. clamps in non-diabetic subjects ( ) and in patients with type 2 diabetes treated with OHA () or insulin (). As previously noted, counter-regulatory failure oc- = curs with increasing duration of type 1 diabetes, despite P 0.025 for non-diabetic versus type 2 diabetic insulin treatment contrast. Reproduced with permission from an initial compensatory catecholamine response [15]. Ref. [23]; © 2002, American Diabetes Association. Does type 2 diabetes follow the same pattern? The glucagon response in type 2 diabetes was diminished Table 21.9 Combined effects of age and type 2 dia- in some studies [17, 38, 39] but preserved in others [36, betes. 37, 40, 41], perhaps reflecting the heterogeneous nature of subjects with type 2 diabetes who would have exhib- 1. Modest attenuation of blood glucose recovery observed ited varying levels of insulin resistance and deficiency (no rise in hepatic glucose production and decline in depending on their body weight, duration of diabetes peripheral utilization) and the underlying genetic component of their con- 2. Some counter-regulatory hormonal responses are re- dition. Most investigators have reported preservation duced, but not epinephrine (adrenaline) of the glucagon response to hypoglycaemia in people 3. Counter-regulatory hormonal response to profound hy- with type 2 diabetes who were most unlikely to be poglycaemia is intact; subtle abnormalities are revealed insulin-deficient at the time of study [36, 37, 41, 49]. by a slow decline in blood glucose By contrast, those subjects with type 2 diabetes who 4. Some tests of cognitive function (psychomotor tests) are were insulin-deficient were shown to have significant more abnormal than in controls impairment of the glucagon response to hypoglycaemia (Figure 21.7) [23, 39]. Growth hormone secretion in response to hypoglycaemia was approximately 50% hypoglycaemia influence the nature of the symptoms. lower compared with non-diabetic subjects [50]. People with type 2 diabetes who were receiving treat- In conclusion, few counter-regulatory hormonal ment with insulin, reported a similar symptom profile deficiencies of significance are present in people associated with hypoglycaemia as a group with type 1 with type 2 diabetes who can be treated with diet diabetes, who had been matched for duration of insulin or oral medication, in contrast to the pronounced therapy, but neither for age nor duration of diabetes counter-regulatory hormonal deficiencies that are [51]. By using the hyperinsulinaemic glucose-clamp exhibited by many individuals with type 1 diabetes technique [41], it could be shown that the hypogly- (Table 21.9). However, this situation changes when caemic symptoms experienced by subjects with type 2 patients with type 2 diabetes progress to a state and type 1 diabetes, who had a similar quality of gly- of pancreatic beta-cell failure, when they then caemic control, were identical. In a different study, the develop the counter-regulatory hormonal deficiencies nature of the symptomatic response to a similar degree characteristic of type 1 diabetes. of hypoglycaemia, induced either by insulin or with tolbutamide, was compared in a group of non-diabetic 21.5.2 Symptoms of hypoglycaemia subjects, and no differences were observed either in the Allowing for differences in age, the symptoms of hypo- nature or in the intensity of symptoms [52]. This would glycaemia do not appear to differ between people with suggest that the agent inducing the hypoglycaemia is type 1 and type 2 diabetes, nor does the agent inducing not important, as identical symptoms were produced 296 CH 21 HYPOGLYCAEMIA when blood glucose was lowered by a similar amount. 21.6 Epidemiology of However, chronic treatment with a particular agent, such as a sulphonylurea, may influence symptomatic hypoglycaemia in elderly responses to hypoglycaemia, and this possibility re- people with diabetes mains to be elucidated. 21.6.1 Determining the frequency of hypoglycaemia 21.5.3 Symptoms in elderly people with type 2 diabetes The frequency of hypoglycaemia in people with diabetes is difficult to determine with accuracy, and many Older people with type 2 diabetes have been shown clinical studies have underestimated the total number to have a lower intensity, and more limited percep- of hypoglycaemic events. In subjects with type 1 dition, of autonomic symptoms of hypoglycaemia than abetes, the retrospective recall of mild (self-treated) age-matched, non-diabetic elderly subjects [38]. In a episodes of hypoglycaemia is inaccurate beyond a pe- descriptive study of 45 elderly patients with type 2 riod of one week, and a prospective recording of hy- diabetes, who were receiving treatment either with poglycaemia is essential to obtain a precise measure insulin or a sulphonylurea, the symptoms of hypo- [54]. The recall of severe hypoglycaemia may be af- glycaemia that were recognized most commonly were fected by amnesia of the event, so that confirmation by non-specific in nature and included weakness, unsteadi- observers and relatives is desirable to verify the accu- ness, sleepiness and faintness [53]. In a retrospective racy of self-reporting. The frequency of hypoglycaemia study of people with type 2 diabetes treated with in- among people with type 2 diabetes is even more diffi- sulin [31], the hypoglycaemia symptoms that were re- cult to ascertain and is prone to underestimation, partly ported with the greatest frequency and intensity were because many are old with memory impairment and a mainly ‘neurological’ in nature, and included unsteadi- limited knowledge of symptoms, and episodes may be ness, light-headedness and poor concentration (see attributed incorrectly to other conditions. Table 21.5). Trembling (71.2%) and sweating (75%) People with type 1 diabetes experience an average also featured prominently [31], contrasting with the of two episodes of symptomatic hypoglycaemia per results of a Canadian study in which it was claimed week, which amounts to thousands of such episodes that the autonomic symptoms of hypoglycaemia in the over a lifetime of treatment with insulin [54]. However, elderly were attenuated [27]. However, the latter study many episodes of mild hypoglycaemia that occur in the did not use an age-specific symptom questionnaire, and community are either unrecognized, or are not docu- differences in symptom questionnaires and in the scor- mented, so that most assessments are likely to signifi- ing methods of inducing hypoglycaemia may account cantly underestimate the magnitude of this problem. By for the differences in symptom profiles that have been contrast, severe hypoglycaemia – defined as an event described. requiring external assistance to recover – occurs less Using the statistical technique of PCA, the hypo- frequently, with an estimated incidence that ranges glycaemia symptoms of elderly people with type 2 from 1.0 to more than 3.0 episodes per patient per year, diabetes could be separated into neuroglycopenic and depending on the duration of the diabetes [55–58]. autonomic groups, but the typical symptoms of a ‘gen- However, although some elderly patients will have an eral malaise’ group of symptoms, such as headache insulin-deficient type 1 diabetes, the majority will have or nausea, were uncommon [31]. However, symptoms type 2 diabetes so it is relevant to examine studies that such as impaired motor coordination and slurring of have considered the latter. speech were prominent. In elderly people, these symp- 21.6.2 Earlier studies of hypoglycaemia in toms may be misinterpreted as representing either cerebral ischaemia, intermittent haemodynamic changes type 2 diabetes associated with cardiac dysrhythmia, or as vasova- The earliest reports on subjects with type 2 diabetes gal and syncopal attacks. Health care professionals who were treated with oral antidiabetic agents should be aware of the age-specific differences in hy- related mainly to sulphonylureas. One Swedish poglycaemic symptoms (see Table 21.1), both from report of the annual incidence of sulphonylurea- the need to identify and treat hypoglycaemia, and for induced hypoglycaemia of sufficient severity to educational purposes. require hospital treatment, recorded a rate of 4.2 per 21.6 EPIDEMIOLOGY OF HYPOGLYCAEMIA IN ELDERLY PEOPLE WITH DIABETES 297 1000 patients [59], but other European surveys have to this area. Unfortunately, the UKPDS reported estimated this to be much lower, at 0.19 to 0.25 per only the prevalence of hypoglycaemia, and missed 1000 patient-years (pt-yr) [60, 61]. This contrasts the opportunity to provide more accurate information with the much higher incidence of insulin-induced on this important complication of diabetes. The hypoglycaemic coma, which has been estimated proportion of patients experiencing hypoglycaemia conservatively at 100 per 1000 pt-yr [62] and severe during the first 10 years of the UKPDS is shown in hypoglycaemia, defined as an episode requiring Table 21.10. Unsurprisingly, people in the intensively external assistance for recovery, is three times more treated group of the UKPDS experienced significantly frequent than coma. A two-year prospective trial more episodes of hypoglycaemia than those in the that involved 321 subjects with type 2 diabetes conventionally treated group [1], but this was still receiving treatment with either chlorpropamide or much lower than estimated frequencies of severe glibenclamide, recorded an incidence of symptomatic hypoglycaemia, ranging from 1.1 to 1.7 episodes per hypoglycaemia of 19 per 1000 pt-yr [63]. Around patient per year in unselected cohorts of people with one-fifth of a relatively young group of 203 patients type 1 diabetes in whom strict glycaemic control was with type 2 diabetes who were receiving treatment not an objective in specialist centres in Denmark [54] with oral sulphonylureas, had experienced symptoms and in Scotland [55], while in a UK multicentre study, suggestive of hypoglycaemia on at least one occasion people with type 1 diabetes of more than 15 years’ during the previous 6 months [64]. Symptoms were duration had an incidence in excess of 3.0 episodes reported most frequently with long-acting preparations per patient per year [58]. such as glibenclamide, and in association with The Veterans Affairs Cooperative Study in the other medications recognized to potentiate their hypoglycaemic effect. USA reported on the frequency of hypoglycaemia in insulin-treated type 2 diabetic patients. In this study, 21.6.3 More recent studies of one group was treated with an intensive insulin regi- hypoglycaemia in type 2 diabetes men, while the other group was administered insulin once daily [65]. The results indicated a higher fre- Interventional studies quency of mild hypoglycaemia in the intensive group With the increasing use of insulin to treat type 2 (intensive versus standard, 16.5 versus 1.5 episodes per diabetes, the findings of a number of studies are patient per annum), but no difference was observed in now available to examine the effects of insulin- the rate of severe hypoglycaemia between the groups induced hypoglycaemia in type 2 diabetes. Many (0.02 episodes per patient per annum). The standard of these studies have also contributed more data treatment group were monitoring blood glucose less on sulphonylurea-induced hypoglycaemia. As few frequently, and the apparently large disparity between large-scale studies have recorded the frequency the groups may be an overestimate. Although these of hypoglycaemic episodes in people with type 2 large interventional trials provide a wealth of informa- diabetes treated with insulin over a protracted period tion on diabetes and its treatments, they commenced of treatment, the UK Prospective Diabetes Study more than a decade ago and do not represent the (UKPDS) offered a valuable opportunity to contribute modern approach to diabetes care, with increasing

Table 21.10 Proportion of patients with type 2 diabetes experiencing hypoglycaemia per year in UK Prospective Diabetes Study over 10 years of the study by principal treatment regimen (mean ﬁgures are shown). Data derived from Ref. [1]. One or more majora episodes Any episode of of hypoglycaemia (%) hypoglycaemia (%) Diet 0.1 1.2 Chlorpropamide 0.4 11.0 Glibenclamide 0.6 17.7 Insulin 2.3 36.5

a Major (severe) hypoglycaemia required third-party help or medical intervention. 298 CH 21 HYPOGLYCAEMIA use of insulin analogues and combination therapy to centre, the incidence of severe hypoglycaemia was as- achieve glycaemic targets. sessed in 600 insulin-treated patients. One in 10 of these patients had type 2 diabetes, and the reported in- cidences of severe hypoglycaemia were 0.73 episodes Observational studies per patient per annum in type 2 diabetes compared to In the USA, a retrospective cohort study was under- 1.7 episodes per patient per annum in type 1 diabetes taken of almost 20 000 elderly people with diabetes [55]. In a study from Tayside, all episodes of severe receiving treatment with either insulin or sulphony- hypoglycaemia that were attended by the emergency lureas, who were enrolling for health insurance [66]. medical services were identified over 12 months [69]. The incidence of fatal hypoglycaemia and of ‘serious During 244 episodes of severe hypoglycaemia, 7.1% hypoglycaemia’ (defined as an emergency admission to had occurred in type 1 diabetes, 7.3% in insulin-treated hospital with a documented blood glucose concentra- type 2 diabetes, and 0.8% in patients taking oral med- tion below 2.8 mmol l−1) was approximately two per ications. In a subsequent study, the same investigators 100 pt-yr. People treated with insulin had a higher in- undertook a short (1-month) prospective study of 267 cidence of ‘serious hypoglycaemia’ than those treated patients with type 1 and insulin-treated type 2 diabetes with sulphonylureas (3 per 100 pt-yr versus 1 per to examine the frequency of hypoglycaemia [70]. The 100 pt-yr). This restricted definition of serious hypo- incidence of severe hypoglycaemia in insulin-treated glycaemia differs considerably from the usual defini- type 2 diabetes was 0.35 episodes per patient per an- tion of severe hypoglycaemia, and so does not provide num, which was one-third of the rate for patients with an accurate picture of the magnitude of the problem. type 1 diabetes (1.15 episodes per patient per annum). The conclusion by the investigators of this study that Similarly, the incidence of all hypoglycaemia (mild sulphonylureas can be used safely in older people, and severe) in patients with insulin-treated type 2 di- without fear of inducing hypoglycaemia, is fallacious abetes was approximately one-third of that affecting as it is not based on accurate measurement, and should those with type 1 diabetes (16.4 versus 42.9 episodes be disregarded. per patient per annum). Furthermore, the patients with In a 6-month questionnaire study from Atlanta of type 2 diabetes were more likely to require treatment African-American patients with type 2 diabetes treated from the emergency services than their counterparts with oral antidiabetic therapy or insulin, 25% had expe- with type 1 diabetes. rienced one episode of hypoglycaemia during the study The definitive study that is relevant to a Western period [67]. Hypoglycaemia increased as treatment was European population with type 2 diabetes is an ob- escalated and severe hypoglycaemia occurred in 0.5% servational prospective multicentre study that was per- of patients, all of whom had been treated with insulin. formed over one year in the UK. Here, the frequency of In a study from Turkey, the frequency of severe hypo- severe hypoglycaemia associated with treatment with glycaemia in 165 patients treated with insulin, most sulphonylureas or with insulin alone (but not with com- of whom had type 2 diabetes, was reported to be bination regimens), was examined in groups of patients very low at 0.15% episodes per patient per year [68], with type 2 diabetes, and compared to two groups of but this figure was derived solely by the retrospective patients with type 1 diabetes of differing duration (<5 examination of hospital case-notes; the significant un- and >15 years) [58]. Patients with HbA1c concentra- derestimation of events associated with this approach tions >9% were excluded, and the mean HbA1c levels makes this data highly dubious. in the study groups ranged from 7.3 to 7.8%. This im- A number of studies from Scotland have reported on portant study revealed that the annual prevalence of the frequency of hypoglycaemia in type 2 diabetes. In severe hypoglycaemia associated with sulphonylureas an early study, the prevalence of severe hypoglycaemia was 7% – a much higher than anticipated figure, and was estimated retrospectively in 104 people with type equivalent to the prevalence of severe hypoglycaemia 2 diabetes of long duration who had progressed to recorded in people with type 2 diabetes who had been pancreatic beta-cell failure and required insulin, and treated with insulin for up to two years. The preva- was not much lower than that of a group of patients lence of severe hypoglycaemia was always lower than with type 1 diabetes who were matched for duration that observed in type 1 diabetes of equivalent du- of insulin therapy but not for age or duration of dia- ration, but progressively increased with the duration betes [51]. In another retrospective study from the same of insulin therapy for type 2 diabetes (Figure 21.8). 21.7 ADVERSE EFFECTS OF HYPOGLYCAEMIA IN THE ELDERLY 299

1.0

0.8

0.6

0.4

0.2 Proportion reporting at least one severe hypo

0.0 type 2 treated with type 2 < 2 yrs type 2 > 5 yrs type 1 < 5 yrs type 1 > 15 yrs sulphonylureas

Figure 21.8 Prevalences of severe hypoglycaemia in patients with types 1 and 2 diabetes treated with different modalities. The proportion of each group experiencing at least one severe self-reported hypoglycaemic episode is shown during 9–12 months of follow-up. The vertical bars indicate 95% conﬁdence intervals. Adapted from Ref. [58].

Because of its prospective design, this study has pro- although this high figure would appear to be an over- vided an accurate measure of the relative risk of devel- estimation compared to the authors’ recent clinical ex- oping severe hypoglycaemia with different treatment perience. In fact, the figure may have declined with modalities and their duration. This has demonstrated the reduced use of long-acting sulphonylureas such as that the prevalence of severe hypoglycaemia associ- chlorpropamide and glibenclamide, that are most likely ated with sulphonylureas is higher than is generally to induce severe hypoglycaemia. appreciated, and that it is also a significant side effect of insulin treatment for type 2 diabetes that becomes 21.7.2 Morbidity greater the longer patients are treated with insulin. The morbidity associated with hypoglycaemia in people with diabetes has been reviewed previously 21.7 Adverse effects of [74–76]. Because of increasing physical frailty and hypoglycaemia in the elderly concomitant diseases such as osteoporosis, the elderly may be more susceptible to physical injury during hypoglycaemia, with fractures of the long bones, 21.7.1 Mortality joint dislocations, soft tissue injuries, head injuries Mortality associated with sulphonylurea-induced hy- and occasionally burns being described as a direct poglycaemia has been calculated as 0.014 to 0.033 consequence of accidents associated with hypogly- per 1000 pt-yr [60, 61], contrasting with an estimated caemia. A trend towards an increase in fracture risk mortality from insulin-induced hypoglycaemia in the in people with diabetes who had a history of previous UK for diabetic patients aged <50 years of approxi- hypoglycaemia was observed in a large case-control mately 0.2 per 1000 pt-yr [71]. In one series, 10% of study in Denmark; this was statistically significant patients with severe sulphonylurea-induced hypogly- for hip fractures, and is thought be the result of caemia who were admitted to hospital subsequently hypoglycaemia-induced falls [77]. Hypothermia may died [72]. Other reviews of the outcome of severe be a direct consequence of hypoglycaemic coma, and hypoglycaemia associated with sulphonylurea therapy the fall in skin temperature during experimentally have cited a mortality rate of approximately 10% [73], induced hypoglycaemia is significantly greater in the 300 CH 21 HYPOGLYCAEMIA presence of the non-selective beta-blocker, propranolol Table 21.11 Potential cardiac sequelae of acute hypo- [78], although this drug is now seldom used. glycaemia. Acute hypoglycaemia provokes a profound Prolongation of QT-interval haemodynamic response secondary to sympathoa- Cardiac arrhythmias drenal activation and the secretion of epinephrine Silent myocardial ischaemia (adrenaline), causing an increase in cardiac output Angina and the workload of the heart [75]. Although this Myocardial infarction degree of haemodynamic stress seldom causes any Cardiac failure pathophysiological problem to the young person with normal cardiac function, in the older individual with Table 21.12 Neuropsychological manifestations of diabetes – who may have underlying macrovascular severe hypoglycaemia. disease – hypoglycaemia may have serious, or even Neurological fatal, consequences. Although the peripheral systolic Focal or generalized convulsions blood pressure rises, the central pressure falls during Coma hypoglycaemia, which may compromise coronary Hemiparesis; transient ischaemic attacks Ataxia; choreoathetosis artery perfusion during diastole and promote my- Focal neurological deficits ocardial ischaemia in people with diabetes of longer Decortication duration, in whom premature stiffening of arteries is common [79]. In diabetic patients who have coronary Psychological heart disease, cardiac arrhythmias may be induced. Cognitive impairment Behavioural/personal changes These have been described during experimentally Automatic or aggressive behaviour; psychosis induced hypoglycaemia and in anecdotal case reports, with atrial fibrillation, nodal rhythms and premature of sensory and motor functions (Table 21.12). Tran- atrial and ventricular contractions all being observed sient ischaemic attacks and transient hemiplegia may during hypoglycaemia in diabetic patients who had be a feature of neuroglycopenia, and less commonly no overt clinical evidence of heart disease [75, 80]. permanent neurological deficits have been described, Sudden death during hypoglycaemia-induced cardiac especially in elderly patients. These are presumably arrhythmia has been described in individual case caused by mechanisms such as direct focal cerebral reports [81, 82]. Transient ventricular tachycardia has damage from glucopenia, acute thrombotic occlusion been observed during experimental hypoglycaemia in secondary to the haemodynamic, haematological and a non-diabetic subject with coronary heart disease, haemorrheological effects of hypoglycaemia, or by while acute myocardial infarction has also been cerebral ischaemia provoked by changes in regional reported in association with acute hypoglycaemia blood flow in the brain [76]. Elderly people who ex- [75]. Acute hypoglycaemia can lengthen the QT perience intermittent hypoglycaemia, particularly from interval on the electrocardiogram, both in non-diabetic the effect of long-acting oral antidiabetic agents, may and diabetic subjects [83]. QT dispersion is a marker be misdiagnosed as having transient ischaemic attacks. of spatial difference in myocardial recovery time In a retrospective review of 778 cases of drug-induced that, when increased, indicates a heightened risk hypoglycaemia, permanent neurological deficit was de- of ventricular arrhythmias and sudden death. This scribed in 5% of the survivors [85]. In a report of was significantly higher during acute insulin-induced 102 cases of hypoglycaemic coma induced either by hypoglycaemia in 13 patients with type 2 diabetes insulin or by glibenclamide, physical injury was re- aged 48–63 years [84]. When combined with the ported in seven patients, myocardial ischaemia in two, effects of catecholamine-mediated hypokalaemia and and stroke in one patient [86]. the profound haemodynamic changes associated with acute hypoglycaemia, the potential for inducing a serious cardiac arrhythmia is enhanced in elderly 21.8 Risk factors for hypoglycaemia people, many of whom have coronary heart disease in the elderly (Table 21.11). Various psychological and neurological manifesta- Retrospective studies have identified advanced tions of acute hypoglycaemia can cause variable loss age and fasting as the two major risk factors 21.8 RISK FACTORS FOR HYPOGLYCAEMIA IN THE ELDERLY 301 associated with sulphonylurea-induced hypoglycaemia elderly Americans, admission to hospital in the preced- [66, 87–90]. However, the increasing number of ing 30 days was identified as the strongest predictor of elderly patients treated with insulin is exposing many severe hypoglycaemia [66]. In those persons aged ≥80 to the traditional risk factors for hypoglycaemia. years, the risk of serious hypoglycaemia was increased These include insulin administration errors, missed further within 30 days of discharge from hospital. meals, increased insulin sensitivity through weight loss, and delayed insulin clearance, often from renal 21.8.1 Alcohol failure. A short prospective study demonstrated that treatment with insulin for longer than 10 years is an Alcohol inhibits hepatic gluconeogenesis, even at important predictor of severe hypoglycaemia in type 2 blood alcohol concentrations that are not usually diabetes [70], while the progressive rise in frequency associated with intoxication, and in people with type of severe hypoglycaemia with increasing duration 1 diabetes it also impairs the ability to perceive of treatment with insulin was illustrated in a large and interpret the symptoms of hypoglycaemia [93]. multicentre study in the UK [58]. Although impaired Small amounts of alcohol can increase the cognitive awareness of hypoglycaemia appears to be uncommon impairment caused by hypoglycaemia in people in insulin-treated type 2 diabetes, it does increase with type 1 diabetes [94], and will presumably the risk of severe hypoglycaemia ninefold [20]. The have a similar effect in people with type 2 diabetes. principal risk factors are listed in Table 21.13, and are In addition, hypoglycaemia may be mistaken as most pertinent to the elderly. inebriation by observers, so delaying the initiation Surveys conducted in Sweden [87, 88] of fatal and of appropriate treatment. Most research into the severe cases of sulphonylurea-induced hypoglycaemia effects of alcohol has been carried out in people with revealed that severe hypoglycaemia was common type 1 diabetes. However, one group performed a during the first month of treatment, was not related to prospective, double-blind, placebo-controlled trial in the dose of the drug used, and that coma and serious 10 older subjects with type 2 diabetes (mean age morbidity were common sequelae. A frequent problem 68 years) to assess the effects of combining alcohol in the elderly is intercurrent illness during which ingestion with fasting [95]. After a 14-h fast, the caloric intake is reduced substantially while the dose of administration of glibenclamide and intravenous sulphonylurea is maintained, so provoking severe hy- alcohol (equivalent to drinking one or two alcoholic poglycaemia. However, adherence to therapy is a com- beverages) resulted in a lower blood glucose nadir − mon problem, particularly with increasing frequency (4.3 ± 1.2 mmol l 1) compared to the group who − of administration and numbers of drugs prescribed [91, did not receive alcohol (5.0 ± 1.4 mmol l 1). In one 92]. Occasionally, hypoglycaemia is induced when subject who developed hypoglycaemia (defined as older people with diabetes are admitted to hospital blood glucose <2.8 mmol l−1 with typical symptoms, and their prescribed dose of oral hypoglycaemic med- or any blood glucose concentration <2.2 mmol l−1) ication is administered accurately, often in the setting during both arms of the study, hypoglycaemia of reduced carbohydrate intake. In a cohort of 20 000 occurred earlier (at 5 h) in the ethanol study compared to the placebo arm (8.5 h). This observation is of Table 21.13 Risk factors for hypoglycaemia in elderly practical importance, since the quantity of alcohol patients with type 2 diabetes mellitus. administered in the study was of a similar amount Age (not dose of drug) to that consumed on a regular basis by many people Impaired renal function with type 2 diabetes. Poor nutrition or fasting Intercurrent illness 21.8.2 Oral antidiabetic agents Duration of diabetes >10 years Alcohol ingestion Hypoglycaemia caused by oral antidiabetic agents Polypharmacy occurs predominantly with insulin secretagogues. Use of long-acting sulphonylureas or insulin Recent hospital admission Although most sulphonylureas can cause fatal hypo- Endocrine deficiency (pituitary, thyroid, adrenal) glycaemia, this has been associated most frequently Deficient counter-regulatory hormonal responses with chlorpropamide and glibenclamide [96]. Impaired awareness of hypoglycaemia Many studies have recorded higher rates of hypoglycaemia with long-acting sulphonylureas such as 302 CH 21 HYPOGLYCAEMIA chlorpropamide and glibenclamide [90, 97, 98]. A Table 21.14 Drug interactions with sulphonylureas community-based study conducted over a 12-year pe- leading to hypoglycaemia. riod in Basle, Switzerland, showed that the treatment of Interaction Drug elderly subjects with type 2 diabetes with longer-acting sulphonylureas was threefold more likely to precipitate Displacement from Aspirin, fibrates, admission to hospital with severe hypoglycaemia than albumin binding sites sulphonamides, warfarin, was the use of short-acting agents [90]. trimethoprim Decreased renal excretion Probenecid, aspirin, Almost all sulphonylurea drugs are metabolized in allopurinol the liver to metabolites that are subsequently excreted Decreased hepatic Warfarin, monoamine in the urine. While most of these metabolites either metabolism oxidase inhibitors have minimal or no metabolic activity, the two ma- Insulin secretagogues Non-steroidal jor hepatic metabolites of chlorpropamide do possess anti-inflammatory drugs; hypoglycaemic activity. In combination with its long low-dose aspirin half-life of approximately 36 h, chlorpropamide is more Inhibition of Alcohol likely to induce prolonged hypoglycaemia in elderly gluconeogenesis Increased peripheral Aspirin people because of the normal age-related decline in glucose uptake the glomerular filtration rate (GFR). Glibenclamide-induced hypoglycaemia may be more pronounced because the drug accumulates within pan- hypoglycaemia were simultaneously taking medica- creatic beta cells, and its metabolites retain some hypo- tions known to increase the risk of hypoglycaemia glycaemic activity. Despite this, many elderly patients [89]. Many drugs can potentiate the effects of with type 2 diabetes who have accompanying risk fac- sulphonylurea agents via a variety of mechanisms; the tors for hypoglycaemia are still receiving these agents [86, 99]. In a review of 150 elderly people with type 2 important drug interactions promoting hypoglycaemia diabetes, 40 of the 45 taking glibenclamide had one are summarized in Table 21.14. or more identifiable risk factors for hypoglycaemia [99]. In general, the use of long-acting sulphonylureas 21.9 Treatment should be avoided in elderly patients with type 2 diabetes, and also in those with renal impairment. Hypoglycaemia is only one of a number of differential However, not all long-acting sulphonylureas pro- diagnoses in an elderly diabetic patient who presents voke hypoglycaemia. Glimepiride is administered once in a comatose state. Although manifestations of cere- daily and stimulates insulin production primarily in re- brovascular disease may be suspected, hypoglycaemia sponse to meals, but the incidence of reported hypogly- should be excluded by measuring blood glucose. caemia is low [100]. In randomized studies comparing However, even when the blood glucose is low, other glimepiride with glibenclamide and gliclazide, the in- common causes such as stroke, intracerebral or sub- cidence of symptomatic hypoglycaemia was lower in arachnoid haemorrhage, head injury and deliberate or glimepiride-treated patients [101]. Furthermore, in a accidental drug or alcohol overdose must not be over- multi-centre, double-blind, controlled trial the modified looked. Failure to respond to treatment with parenteral release (MR) preparation of gliclazide was reported glucose should immediately arouse suspicion that there to cause less hypoglycaemia compared to glimepiride might be another cause for the coma. Failure to recover (3.7% and 8.9%, respectively). Importantly, in this lat- consciousness following an episode of severe hypogly- ter study more than one-third of the subjects were aged caemia may be associated with cerebral oedema, which >65 years [102]. has a poor prognosis. Affected patients require management in hospital in an intensive care unit, with the 21.8.3 Adverse drug interactions use of agents such as mannitol, steroids and high-flow Several adverse drug interactions between sulphony- oxygen. This suspected presence of cerebral oedema lureas and other commonly prescribed medications must be confirmed, and other causes of coma excluded, are recognized that increase the risk of sulphonylurea- by neuroimaging of the brain. Although most patients induced hypoglycaemia. In a comprehensive re- recover rapidly from an episode of hypoglycaemia, it view, 15% of patients with sulphonylurea-induced can take up to an hour after blood glucose has been 21.9 TREATMENT 303

Hypoglycaemia

Type 1 Type 2 diabetes diabetes

Patient Patient drowsy/ Insulin-treated Sulphonylurea conscious unconscious

Oral Oral Patient Patient drowsy/ Patient Patient drowsy/ carbohydrate carbohydrate; conscious unconscious conscious unconscious (>20 g, iv 20% liquid/solid) dextrose (50 ml) Oral Oral iv dextrose glucagon Oral 50 ml 20% ± iv carbohydrate carbohydrate; carbohydrate (1mg) im/iv (>20 g, (>20 g, infusion of iv 20% 10% dextrose liquid /solid) dextrose liquid/solid) (50 ml) glucagon consider adjunctive (1mg) im/iv therapies: • diaxoxide, • octreotide

Figure 21.9 Treatment measures for acute hypoglycaemia. restored to normal for all cognitive functions to be prove to be ineffective in patients with protracted hy- fully recovered [103–105]. Recovery may take longer poglycaemia whose stores of glycogen are exhausted, in the elderly, although this point has not been studied in people with advanced liver disease or alcohol abuse, in older people. and in those with malnutrition or inanition. The treatment of hypoglycaemia in people with type 2 diabetes follows the same basic principles as for type 21.9.1 Sulphonylurea-induced 1 diabetes, but important differences are present be- hypoglycaemia tween those treated with insulin and those treated with Mild sulphonylurea-induced hypoglycaemia is treated sulphonylureas (Figure 21.9). In the conscious individ- in a similar fashion to insulin-induced hypoglycaemia, ual, acute hypoglycaemia is treated with rapid-acting by the ingestion of rapid-acting glucose followed oral carbohydrate, usually in the form of glucose tablets by a longer-acting complex carbohydrate in the or confectionery such as sweets or chocolate. Bev- form of bread, biscuits, cereal or other alternatives. erages with a high glucose content are also suitable, Sulphonylurea-induced hypoglycaemic coma requires such as fresh orange juice. Following ingestion, some inpatient management and, following the adminis- form of long-acting carbohydrate should be consumed tration of intravenous dextrose, the patient should to prevent recurrence of the hypoglycaemia. In the not be discharged immediately from hospital, despite apparent recovery, as hypoglycaemia from this cause drowsy or unconscious patient who cannot swallow, an is often associated with relapse and is therefore intravenous injection of dextrose (50 ml of a 20% so- protracted [96, 106, 107]. Glucagon can stimulate lution) will cause a rapid reversal of neuroglycopenia. insulin secretion in people with type 2 diabetes who Glucagon is very effective if administered to people have residual pancreatic beta-cell function [108], and treated with insulin that are semi-conscious or in a is therefore said to be contraindicated in the treatment hypoglycaemic coma. Glucagon (1 mg, intramuscular of sulphonylurea-induced hypoglycaemia, although injection) can be administered either by a friend or rela- the evidence for this supposition is very limited. Fol- tive who is familiar with the technique or by paramed- lowing a bolus intravenous injection of 50 ml of 20% ical staff, but it may induce nausea and/or vomiting. dextrose, many patients require prolonged intravenous It acts through promoting hepatic glycogenolysis to infusion of 10% (or even 20%) dextrose to maintain stimulate hepatic glucose output, so on occasion may a blood glucose concentration above 5.0 mmol l−1. 304 CH 21 HYPOGLYCAEMIA The duration of the intravenous infusion will depend only 32% of those treated with insulin denied any upon the half-life of the sulphonylurea ingested, and knowledge of hypoglycaemia [53]. This lack of knowl- in cases of drug overdose it may be necessary to edge extends to the treatment of hypoglycaemia, with continue the dextrose infusion for several days. around 20% of elderly people with insulin-treated diabetes being unfamiliar with the appropriate measures 21.9.2 Adjunctive therapies required to treat symptomatic hypoglycaemia [114]. This reflects the problems of providing information Diazoxide has a direct inhibitory effect on insulin se- about side effects of therapy and the putative defi- cretion, and has been used as an adjunct to dextrose ciencies of the education for an elderly group with infusion in the treatment of sulphonylurea-induced hy- type 2 diabetes, some of whom have memory impair- poglycaemia [109, 110]. In the unconscious patient, ment and may have difficulty retaining information diazoxide can be infused intravenously (300 mg over about the risk of drug-induced hypoglycaemia. Hence, 30 min, repeated 4-hourly if necessary), or in the con- regular reinforcement is required regarding potential scious patient can be administered orally (300 mg every symptoms and the treatment of hypoglycaemia [115]. 4 h). Unfortunately, this is unlikely to be a practi- Most elderly patients with type 2 diabetes are treated cal option because this drug is no longer commer- in primary care, and education about therapeutic side cially available and is now very difficult to obtain. effects must not be overlooked, particularly on the Octreotide, a long-acting synthetic analogue of somato- mistaken assumption that hypoglycaemia is rarely a statin, was used to treat sulphonylurea-induced hypo- clinical problem. glycaemia successfully in a non-diabetic patient who had taken an overdose of tolbutamide [111]. Octreotide reverses the hyperinsulinaemia induced by excessive 21.11 Prevention of hypoglycaemia administration of a sulphonylurea and, in contrast to diazoxide, reduces the amount of dextrose required to All patients taking insulin or sulphonylureas should re- maintain euglycaemia [112]. ceive education about the symptoms of hypoglycaemia and the most appropriate form of self-treatment, and 21.10 Knowledge of symptoms should be advised to eat regular meals so as to ensure an adequate intake of carbohydrate. All patients taking Many patients and their relatives are remarkably un- insulin should carry a card or other form of identifica- informed about the symptoms of hypoglycaemia [53, tion, stating that they have diabetes and what treatment 113] and its emergency treatment. In one study, 9% they are taking. People who are self-administering in- of elderly people who were taking sulphonylureas or sulin should carry a source of rapid-acting carbohy- insulin were unable to state any of the symptoms as- drate. Individuals at increased risk of hypoglycaemia sociated with hypoglycaemia. Only 20% of the group should be identified, such as those aged >70 years and said that they had been told about the symptoms of hy- people with renal impairment. In the elderly type 2 di- poglycaemia. Of particular concern, people who were abetic, the longer-acting sulphonylurea agents such as living alone had no better knowledge of symptoms, chlorpropamide and glibenclamide should be avoided, despite being at greater risk from the consequences of and short-acting sulphonylureas such as glipizide and hypoglycaemia [113]. Knowledge of the symptoms of gliclazide used and commenced at the lowest dose, hypoglycaemia was unrelated to the duration of dia- titrated against the patient’s fasting or postprandial betes and living alone, and did not differ between men blood glucose level. Persistently low blood glucose and women. However, older people knew even less values will require a reduction in dose. about hypoglycaemia than younger individuals with Consideration should be given to the use of alter- diabetes. Many people with diabetes taking sulphony- native oral antidiabetic agents that either do not cause lureas did not know about the potential risk of develop- or have a much lower risk of hypoglycaemia, such as ing hypoglycaemia, whereas people treated with insulin metformin and the thiazolidinediones, and the newer were more knowledgeable on this subject. These find- incretin mimetics such as the gliptins, or the injectable ings were confirmed by a different study, in which GLP-1 agonist, exenatide. The utilization of fast-acting 88% of elderly people taking sulphonylureas did not insulin analogues and fixed mixtures containing these know the potential risk of hypoglycaemia, whereas insulin preparations may also be of value in elderly 21.12 SUMMARY AND CONCLUSIONS 305

Table 21.15 Prevention of hypoglycaemia. • Although most elderly people have type 2 diabetes, insulin therapy is increasingly being used. Although 1. Identify high-risk patients: advanced age, renal impair- the counter-regulatory hormonal deﬁciencies and im- ment paired awareness of hypoglycaemia that are common 2. Avoid long-acting sulphonylurea preparations in people with type 1 diabetes are observed infre- 3. Use short-acting sulphonylurea (e.g. gliclazide) quently in elderly people with insulin-treated dia- 4. Be aware of drug interactions betes, these acquired hypoglycaemia syndromes may increase in prevalence. 5. Educate patients and relatives/carers about hypoglycaemia • Although hypoglycaemia is a common side 6. Consider agents that do not cause hypoglycaemia effect of insulin therapy, it occasionally may be a life-threatening problem in people taking long-acting sulphonylureas, especially those patients people treated with insulin, in particular to avoid noc- of advanced years who have concomitant renal turnal hypoglycaemia. Any oral agent may promote impairment and/or inadequate caloric intake. hypoglycaemia in situations where the patient is un- • Long-acting sulphonylureas, such as glibenclamide able to eat normally, and the dose may have to be and chlorpropamide, should be avoided in the el- reduced, or the drug temporarily discontinued. derly. Adverse drug interactions with sulphonylurea In order to improve adherence to a prescribed oral agents can cause profound hypoglycaemia. antidiabetic therapy, a dosette box is often used in el- • derly patients, although this might also reduce the risk If the treatment of hypoglycaemic coma with in- of overdose due to forgetfulness and to the difﬁcul- travenous dextrose is unsuccessful, other causes of ties of coping with polypharmacy. Clinicians must be coma should be sought urgently, and the presence of particularly aware of drug interactions with sulphony- cerebral oedema excluded by neuroimaging. lureas, and anticipate and modify their prescribing pat- • People with diabetes who are treated with sulphony- terns accordingly (see Table 21.14). It is important to lureas or insulin, and their relatives or carers, must be provide adequate information to relatives and other car- fully informed of the symptoms and effects of hy- ers who may be administering medications and moni- poglycaemia and its emergency management. This toring blood glucose at home, because they may lack should be reinforced with appropriate education at knowledge of the symptoms and consequences of hy- intervals. poglycaemia (Table 21.15).

21.12 Summary and conclusions References

To summarize: 1. UK Prospective Diabetes Study Group (1998) Inten- sive blood-glucose control with sulphonylureas or in- • Hypoglycaemia in the elderly person with type 2 sulin compared with conventional treatment and risk diabetes is a potentially serious and underestimated of complications in patients with type 2 diabetes clinical problem that has a signiﬁcant morbidity and (UKPDS 33). Lancet, 352, 837–53. mortality. 2. Cryer PE (1993) Glucose counterregulation: prevention and correction of hypoglycemia in humans. Am • Many symptoms of hypoglycaemia in the elderly J Physiol, 264, E149–55. are ‘neurological’ in type, the cause of which may 3. Gerich JE (1988) Glucose counterregulation and its be misinterpreted as representing a cerebrovascular impact on diabetes mellitus. Diabetes, 37, 1608–17. event or cardiac arrhythmia. 4. Cryer PE, Binder C, Bolli GB, Cherrington AD, Gale EAM, Gerich JE and Sherwin RS (1989) Hy- • The effects of ageing and of type 2 diabetes on the poglycemia in IDDM. Diabetes, 38, 1193–9. hormonal counter-regulation, symptoms and aware- 5. Towler DA, Havlin CE, Craft S and Cryer P (1993) ness of hypoglycaemia should be considered when Mechanism of awareness of hypoglycemia. Perception deciding upon targets for glycaemic control in el- of neurogenic (predominantly cholinergic) rather than derly people with diabetes. neuroglycopenic symptoms. Diabetes, 42, 1792–8. 306 CH 21 HYPOGLYCAEMIA

6. Corrall RJM, Frier BM, Davidson NMcD, Hop- causal relation with diabetic autonomic neuropathy. kins WM and French EB (1983) Cholinergic Br Med J, 301, 783–7. manifestations of the acute autonomic reaction to 19. Cryer PE (2002) Hypoglycaemia: the limiting factor hypoglycaemia in man Clin Sci, 64, 49–53. in the management of type 1 and type 2 diabetes. 7. Macdonald IA and Maggs DA (1993) Cutaneous Diabetologia, 45, 937–48. blood flow, sweating, tremor and temperature regu- 20. Henderson JN, Allen KV, Deary IJ and Frier BM lation in hypoglycaemia. In: Hypoglycaemia and di- (2003) Hypoglycaemia in insulin-treated type 2 dia- abetes: Clinical and physiological aspects,BMFrier betes: frequency, symptoms and impaired awareness. and BM Fisher (eds), Edward Arnold, London, pp. Diabet Med, 20, 1016–21. 132–43. 21. Cryer PE, Fisher JN and Shamoon H (1994) Hypo- 8. Deary IJ, Hepburn DA, MacLeod KM and Frier BM glycemia. Diabetes Care, 17, 734–55. (1993) Partitioning the symptoms of hypoglycaemia 22. Cryer PE (1992) Iatrogenic hypoglycemia as a cause using multi-sample confirmatory factor analysis. Dia- of hypoglycemia-associated autonomic failure in betologia, 36, 771–7. IDDM. A vicious cycle. Diabetes, 41, 255–60. 9. Frier BM and Fisher BM (eds) (1993) Hypoglycaemia 23. Segel SA, Paramore DA and Cryer PE (2002) in Diabetes: Clinical and Physiological Aspects,Ed- Hypoglycemia-associated autonomic failure in ward Arnold (Publishers) Ltd, London. advanced type 2 diabetes. Diabetes, 51, 724–33. 10. Deary, I.J. (2007) Symptoms of hypoglycaemia and 24. Cryer PE (2004) Diverse causes of hypoglycemia- effects on mental performance and emotions. In: BM associated autonomic failure in diabetes. N Engl, 350, Frier and BM Fisher (eds) Hypoglycaemia in Clinical 2272–9. Diabetes, 2nd edition. John Wiley & Sons, Chichester, 25. Marker JC, Cryer PE and Clutter WE (1992) pp. 25–48. Attenuated glucose recovery from hypoglycemia in 11. McAulay V, Deary IJ, Frier BM (2001) Symptoms of the elderly. Diabetes, 41, 671–8. hypoglycaemia in people with diabetes Diabet Med, 26. Meneilly GS, Minaker KL, Young JB, Landsberg L 18, 690–705. and Rowe JW (1985) Counterregulatory responses to 12. Vea H, Jorde R, Sager G, Vaaler S and Sundsfjord J insulin-induced glucose reduction in the elderly. J (1992) Reproducibility of glycaemic thresholds for ac- Clin Endocrinol Metab, 61, 178–82. tivation of counterregulatory hormones and hypogly- 27. Meneilly GS, Cheung E and Tuokko H (1994) Altered caemic symptoms in healthy subjects. Diabetologia, responses to hypoglycemia of healthy elderly people. 35, 958–61. J Clin Endocrinol Metab, 78, 1341–8. 13. Frier BM (2007) Impaired awareness of hypogly- 28. Ortiz-Alonso FJ, Galecki A, Herman WH, Smith caemia. In: Hypoglycaemia in Clinical Diabetes, 2nd MJ, Jacquez JA and Halter JB (1994) Hypoglycemia edition. BM Frier and M Fisher (eds), John Wiley & counterregulation in elderly humans: relationship to Sons, Chichester, pp. 141–70. glucose levels. Am J Physiol, 267, E497–506. 14. Amiel SA (2007) Risks of strict glycaemic control. 29. Brierley EJ, Broughton DL, James OWF and Alberti In: Hypoglycaemia in Clinical Diabetes, 2nd edition. KGMM (1995) Reduced awareness of hypoglycaemia BM Frier and M Fisher (eds), John Wiley & Sons, in the elderly despite an intact counterregulatory re- Chichester, pp. 171–89. sponse. Quart J Med 88, 439–45. 15. Gerich JE and Bolli GB (1993) Counterregulatory 30. Matyka K, Evans M, Lomas J, Cranston I, Macdonald failure. In: Hypoglycaemia and Diabetes: Clinical and I and Amiel SA (1997) Altered hierarchy of protec- Physiological Aspects. BM Frier and BM Fisher (eds), tive responses against severe hypoglycemia in normal Edward Arnold, London, pp. 253–67. ageing in healthy men. Diabetes Care, 20, 135–41. 16. White NH, Skor DA, Cryer PE, Levandoski LA, 31. Jaap AJ, Jones GC, McCrimmon RJ, Deary IJ and Bier DM and Santiago JV (1983) Identification of Frier BM (1998) Perceived symptoms of hypogly- type 1 diabetic patients at increased risk for hypo- caemia in elderly type 2 diabetic patients treated with glycemia during intensive therapy. New Engl J Med, insulin. Diabet Med, 15, 398–401. 308, 485–91. 32. Reynolds C, Molnar GD, Horwitz DL, Rubenstein 17. Bolli GB, Tsalikian E, Haymond MW, Cryer PE and AH, Taylor WF and Jiang NS (1977) Abnormalities of Gerich JE (1984) Defective glucose counterregulation endogenous glucagon and insulin in unstable diabetes. after subcutaneous insulin in non-insulin-dependent Diabetes, 26, 36–45. diabetes mellitus. J Clin Invest, 73, 1532–41. 33. Nonaka K, Toyoshima T, Yoshida T, Matsuyama T, 18. Ryder RE, Owens DR, Hayes TM, Ghatei MA and Trarni S and Nishikawa M (1977) The nature of hy- Bloom SR (1990) Unawareness of hypoglycaemia perglucagonaemia in diabetes mellitus. In: Glucagon: and inadequate hypoglycaemic counterregulation: no its Role in Physiology and Clinical Medicine,TPFoa, 21.12 SUMMARY AND CONCLUSIONS 307

JS Jajos, NL Foa (eds), Springer-Verlag, New York, 45. Diamond MP, Hallerman L, Starick-Zych K, Jones pp. 662–77. TW, Connolly-Howard M, Tamborlane WV and Sher- 34. Polonsky KS, Herold KC, Gilden JL, Bergenstal RM, win RS (1991) Suppression of counterregulatory hor- Fang VS, Moossa AR and Jaspan JB (1984) Glucose mone response to hypoglycemia in normal man. J Clin counterregulation in patients after pancreatectomy. Endocrinol Metab, 73, 1123–8. Comparison with other clinical forms of diabetes. 46. Liu D, Moberg E, Kollind M, Lins PE and Adamson Diabetes, 33, 1112–19. U (1991) A high concentration of circulating insulin 35. Levitt NS, Vinik AI, Sive AA, Child P and Jackson suppresses the glucagon response to hypoglycemia in WPU (1979) Studies on plasma glucagon concentra- normal man. J Clin Endocrinol Metab, 73, 1123–8. tion in maturity-onset diabetics with autonomic neu- 47. Davis SN, Goldstein RE, Jacobs J, Price L, Wolfe R ropathy. Diabetes, 28, 1015–21. and Cherrington AD (1993) The effects of differing insulin levels on the hormonal and metabolic response 36. Boden G, Soriano M, Hoeldtke RD and Owen to equivalent hypoglycemia in normal humans. Dia- OE (1983) Counterregulatory hormone releases betes, 42, 263–72. and glucose recovery after hypoglycemia in 48. Fanelli C, Pampanelli S, Epifano L, Rambotti AM, non-insulin-dependent diabetic patients. Diabetes, Ciofetta M, Modarelli F, et al. (1994) Relative roles of 32, 1055–9. insulin and hypoglycaemia on induction of neuroen- 37. Heller SR, Macdonald IA and Tattersall RB (1987) docrine responses to, symptoms of, and deterioration Counterregulation in type 2 (non-insulin-dependent) of cognitive function in hypoglycaemia in male and diabetes mellitus. Normal endocrine and glycaemic re- female humans. Diabetologia, 37, 797–807. sponses, up to ten years after diagnosis. Diabetologia, 49. Spyer G, Hattersley A, Macdonald IA, Amiel S and 30, 924–9. MacLeod KM (2000) Hypoglycaemic counterregula- 38. Meneilly GS, Cheung E and Tuokko H (1994) Coun- tion at normal blood glucose concentrations in patients terregulatory hormone responses to hypoglycemia with well controlled type 2 diabetes. Lancet, 356, in the elderly patient with diabetes. Diabetes, 43, 1970–4. 403–10. 50. Israelian Z, Szoke E, Woerle J, Bokhari S, Schorr 39. Shamoon H, Friedman S, Canton C, Zacharowicz L, M, Schwenke DC, Cryer PE, Gerich JE and Meyer Hu M and Rossetti L (1994) Increased epinephrine C (2006) Multiple defects in counterregulation of and skeletal muscle responses to hypoglycemia in hypoglycemia in modestly advanced type 2 diabetes non-insulin-dependent diabetes mellitus. J Clin Invest, mellitus. Metab Clin Exp, 55, 593–8. 93, 2562–71. 51. Hepburn DA, MacLeod KM, Pell ACH, Scougal IJ 40. Korzon-Burakowska A, Hopkins D, Matyka K, Lomas and Frier BM (1993) Frequency and symptoms of J, Pernet A, Macdonald I and Amiel S (1998) Effects hypoglycaemia experienced by patients with type 2 of glycemic control on protective responses against diabetes treated with insulin. Diabet Med, 10, 231–7. hypoglycemia in type 2 diabetes. Diabetes Care, 21, 52. Peacey SR, George E, Rostami-Hodjegan A, Bedford 283–90. C, Harris N, Hardisty CA, Tucker GT, Macdonald 41. Levy CJ, Kinsley BT, Bajaj M and Simonson DC IA and Heller SR (1996) Similar physiological and symptomatic responses to sulphonylurea and insulin (1998) Effect of glycemic control on glucose counter- induced hypoglycaemia in normal subjects. Diabet regulation during hypoglycemia in NIDDM. Diabetes Med, 13, 634–41. Care, 21, 1330–8. 53. Thomson FJ, Masson EA, Leeming JT and Boulton 42. Amiel SA, Maran A, Powrie JK, Umpleby AM and AJM (1991) Lack of knowledge of symptoms of hy- Macdonald IA (1993) Gender differences in coun- poglycaemia by elderly diabetic patients. Age Ageing, terregulation to hypoglycaemia. Diabetologia, 36, 20, 404–6. 460–4. 54. Pramming S, Thorsteinsson B, Bendtson I and Binder 43. Davis SN, Fowler S and Costa F (2000) Hypo- C (1991) Symptomatic hypoglycaemia in 411 type 1 glycemic counterregulatory responses differ between diabetic patients. Diabet Med, 8, 217–22. men and women with type 1 diabetes. Diabetes, 49, 55. MacLeod KM, Hepburn DA and Frier BM (1993) 56–72. Frequency and morbidity of severe hypoglycaemia 44. Sandoval DA, Ertl AC, Richardson MA, Tate DB and in insulin-treated diabetic patients. Diabet Med, 10, Davis SN (2003) Estrogen blunts neuroendocrine and 238–45. metabolic responses to hypoglycemia. Diabetes, 52, 56. ter Braak EWMT, Appleman AMMMF, van de Laak 1749–55. M, Stolk RP, van Haeften TW and Erkelens DW 308 CH 21 HYPOGLYCAEMIA

(2000) Clinical characteristics of type 1 diabetic pa- DARTS/MEMO Collaboration (2003) Frequency of tients with and without severe hypoglycemia. Dia- severe hypoglycemia requiring emergency treatment betes Care, 23, 1467–71. in type 1 and type 2 diabetes: a population-based 57. Pedersen-Bjergaard U, Pramming S, Heller SR, Wal- study of health service resource use. Diabetes Care, lace TM, Rasmussen AK, Jorgensen HV, Matthews 26, 1176–80. DR, Hougaard P and Thornsteinsson B (2004) Se- 70. Donnelly LA, Morris AD, Frier BM, Ellis JD, Donnan vere hypoglycaemia in 1076 adult patients with type PT, Durrant R et al., for the DARTS/MEMO Collab- 1 diabetes: influence of risk markers and selection. oration (2005) Frequency and predictors of hypogly- Diabetes Metab Res Rev, 20, 479–86. caemia in type 1 and insulin-treated type 2 diabetes: 58. UK Hypoglycaemia Study Group (2007) Risk of hy- a population-based study. Diabet Med 22, 749–55. poglycaemia in types 1 and 2 diabetes: effects of 71. Tunbridge WMG (1981) Factors contributing to treatment modalities and their duration. Diabetologia, deaths of diabetics under fifty years of age. Lancet, 50, 1140–7. ii, 569–72. 59. Dahlen M, Bergman U, Idman L, Martinsson L and 72. Seltzer HS (1972) Drug-induced hypoglycemia: a Karlsson G (1984) Epidemiology of hypoglycaemia review based on 473 cases. Diabetes, 21, 955–66. in patients on oral antidiabetic drugs in the island of 73. Campbell IW (1993) Hypoglycaemia and Type II di- Gotland, Sweden. Acta Endocrinol, 263 (Suppl. 21), abetes: sulphonylureas. In: Hypoglycaemia and dia- Abstract. betes. Clinical and physiological aspects,BMFrier 60. Berger W (1985) Incidence of severe side-effects and BM Fisher (eds), Edward Arnold, London, pp. during therapy with sulfonylureas and biguanides. 387–92. Horm Metab Res, 15 (Suppl.), 111–15. 74. Frier BM (1992) Hypoglycaemia - how much harm? 61. Campbell IW (1985) Metformin and the sulphony- Hospital Update, 18, 876–84. lureas: the comparative risk. Horm Metab Res 15, 75. Fisher BM and Heller SR (2007) Mortality, cardiovas- 105–11. cular morbidity and possible effects of hypoglycaemia 62. Gerich JE (1989) Oral hypoglycemic agents. New on diabetic complications. In: Hypoglycaemia in Clin- Engl J Med, 321, 1231–45. ical Diabetes, 2nd edition, BM Frier and M Fisher 63. Clarke BF and Campbell IW (1975) Long-term com- (eds), John Wiley & Sons, Chichester, pp. 265–83. parative trial of glibenclamide and chlorpropamide in 76. Perros P and Deary IJ (2007) Long-term effects of diet-failed, maturity onset diabetics. Lancet, i, 246–8. hypoglycaemia on cognitive function and the brain in 64. Jennings AM, Wilson RM and Ward JD (1989) Symp- diabetes. In: Hypoglycaemia in Clinical Diabetes, 2nd tomatic hypoglycemia in NIDDM patients treated edition, BM Frier and M Fisher (eds), John Wiley & with oral hypoglycemic agents. Diabetes Care, 12, 203–8. Sons, Chichester, pp. 285–307. 65. Abraira C, Colwell JA, Nuttall FQ, Sawin CT, Nagel 77. Vestergaard P, Rejnmark L and Mosekilde L (2005) NJ, Comstock JP, Emanuele NV, Levin SR, Hender- Relative fracture risk in patients with diabetes mel- son W and Lee HS (1995) Veterans Affairs Coopera- litus, and the impact of insulin and oral antidiabetic tive Study on glycemic control and complications in medication on relative fracture risk. Diabetologia, 48, type II diabetes mellitus (VACSDM). Results of the 1292–9. feasibility trial. Veterans Affairs Cooperative Study in 78. Macdonald IA, Bennett T, Gale EAM, Green JH and Type II Diabetes. Diabetes Care, 18, 1113–23. Walford S (1982) The effect of propranolol or meto- 66. Shorr RI, Ray WA, Daugherty JR and Griffin MR prolol on thermoregulation during insulin-induced hy- (1997) Incidence and risk factors for serious hypo- poglycaemia in man. Clin Sci, 63, 301–10. glycemia in older persons using insulin or sulfony- 79. Sommerfield AJ, Wilkinson IB, Webb DJ and Frier lureas. Arch Intern Med, 157, 1681–6. BM (2007) Vessel wall stiffness in type 1 diabetes 67. Miller CD, Philips LS, Ziemer DC, Gallina DL, Cook and the central hemodynamic effects of acute hy- CB and El-Kebbi IM (2001) Hypoglycemia in patients poglycemia. Am J Physiol Endocrinol Metab, 293, with type 2 diabetes mellitus. Arch Intern Med, 161, E1274–9. 1653–59. 80. Lindstrom¨ T, Jorfeldt L, Tegler L and Arnqvist HJ 68. Gurlek A, Erbas T and Gedik O (1999) Frequency (1992) Hypoglycaemia and cardiac arrhythmias in of severe hypoglycaemia in type 1 and type 2 dia- patients with type 2 diabetes mellitus. Diabet Med, betes during conventional insulin therapy. Exp Clin 9, 536–41. Endocrinol Diabetes, 107, 220–4. 81. Frier BM, Barr StCG and Walker JD (1995) Fatal 69. Leese GP, Wang J, Broomhall J, Kelly P, Mars- cardiac arrest following acute hypoglycaemia in a den A, Morrison W, Frier BM, Morris AD, the diabetic patient. Pract Diab Internat, 12, 284. 21.12 SUMMARY AND CONCLUSIONS 309

82. Burke BJ and Kearney TK (1999) Hypoglycaemia and 97. Brodows RG (1992) Benefits and risks with glyburide cardiac arrest. Pract Diab Internat, 16, 189–90. and glipizide in elderly NIDDM patients. Diabetes 83. Marques JLB, George E, Peacey SR, Harris ND, Mac- Care, 15, 75–80. donald IA, Cochrane T and Heller SR (1997) Altered 98. Tessier D, Dawson K, Tetrault JP, Bravo G and ventricular repolarisation during hypoglycaemia in pa- Meneilly GS (1994) Glibenclamide vs gliclazide in tients with diabetes. Diabet Med, 14, 648–54. type 2 diabetes of the elderly. Diabet Med, 11, 84. Landstedt-Hallin L, Englund A, Adamson U and Lins 974–80. PE (1999) Increased QT dispersion during hypogly- 99. Yap WS, Peterson GM, Vial JH, Randall CT and caemia in patients with type 2 diabetes. J Intern Med, Greenaway TM (1998) Review of management of 246: 299–307. type 2 diabetes mellitus. J Clin Pharm Ther, 23, 85. Seltzer HS (1979) Severe drug-induced hypogly- 457–65. caemia: a review. Compr Ther, 5, 21–9. 100. Campbell RK (1998) Glimepiride: Role of a new sul- 86. Ben-Ami H, Nagachandran P, Mendelson A and Ed- fonylurea in the treatment of type 2 diabetes mellitus. oute Y (1999) Drug-induced hypoglycemic coma in Ann Pharmacother, 32, 1044–52. 102 diabetic patients. Arch Intern Med, 159, 281–4. 101. Schneider, J (1996) An overview of the safety and tol- 87. Asplund K, Wilholm BE and Lithner F (1983) Gliben- erance of glimepiride. Horm Metab Res, 28: 413–18. clamide associated hypoglycaemia: a report on 57 102. Schernthaner G, Grimaldi A, Di Mario U, Drzewoski cases. Diabetologia, 24, 412–17. J, Kempler P, Kvapil M, Shaw KM, the GUIDE 88. Asplund K, Wilholm BE and Lundman B (1991) Se- Study (2004) Double blind comparison of once-daily vere hypoglycaemia during treatment with glipizide. gliclazide MR and glimepiride in type 2 diabetic Diabet Med, 8, 726–31. patients. Eur J Clin Invest, 34, 535–42. 89. Seltzer HS (1989) Drug-induced hypoglycemia: a 103. Lindgren M, Eckert B, Stenberg G and Agardh CD (1996) Restitution of neurophysiological functions, review of 1418 cases. Endocrinol Metab Clin North performance, and subjective symptoms after moderate Am, 18, 163–83. insulin-induced hypoglycaemia in non-diabetic men. 90. Stahl M and Berger W (1999) Higher incidence of Diabet Med, 13, 218–25. severe hypoglycaemia leading to hospital admission 104. Evans ML, Pernet A, Lomas J, Jones J and Amiel SA in type 2 diabetic patients treated with long-acting (2000) Delay in onset of awareness of acute hypo- versus short-acting sulphonylureas. Diabet Med, 16, glycemia and of restoration of cognitive performance 586–90. during recovery. Diabetes Care, 23, 893–7. 91. Paes AHP, Bakker A and Soe-Agnie CJ (1997) Impact 105. Fanelli CG, Pampanelli S, Porcellati F et al. (2003) of dosage frequency on patient compliance. Diabetes Rate of fall of blood glucose and physiological 20 Care, , 1512–17. responses of counterregulatory hormones, clinical 92. Donnan PT, MacDonald TM and Morris AD (2002) symptoms and cognitive function to hypoglycaemia Adherence to prescribed oral hypoglycaemic medica- in type 1 diabetes mellitus in the postprandial state. tion in a population of patients with type 2 diabetes: a Diabetologia, 46, 53–64. retrospective cohort study. Diabet Med, 19, 279–84. 106. Gale EAM (1980) Hypoglycaemia. Clin Endocrinol 93. Kerr D, Macdonald IA, Heller SR and Tattersall RB Metab, 9, 461–75. (1990) Alcohol causes hypoglycaemic unawareness 107. Marks V (1981) Drug-induced hypoglycaemia. In: in healthy volunteers and patients with type I Hypoglycaemia, 2nd edition, V Marks and FC Rose (insulin-dependent) diabetes. Diabetologia, 33, (eds), Blackwell Scientific Publications, Oxford, pp. 216–21. 357–86. 94. Cheyne EH, Sherwin RS, Lunt MJ, Cavan DA, 108. Marri G, Cozzolino G and Palumbo R (1968) Thomas PW and Kerr D (2004) Influence of alco- Glucagon in sulphonylurea hypoglycaemia? Lancet, hol on cognitive performance during mild hypogly- i, 303–4. caemia: implications for type 1 diabetes. Diabet Med, 109. Johnson SF, Schade DS and Peake GT (1977) 21, 230–7. Chlorpropamide-induced hypoglycemia: successful 95. Burge MR, Zeise TM, Sobhy TA, Rassam AG and treatment with diazoxide. Am J Med, 63, 799–804. Schade DS (1999) Low-dose ethanol predisposes 110. Palatnick W, Meatherall RC and Tenenbein M (1991) elderly fasted patients with type 2 diabetes to Clinical spectrum of sulfonylurea overdose and expe- sulfonylurea-induced low blood glucose. Diabetes rience with diazoxide therapy. Arch Intern Med, 151, Care, 22, 2037–43. 1859–62. 96. Ferner RE and Neil HA (1988) Sulphonylureas and 111. Krentz AJ, Boyle PJ, Justice KM, Wright AD and hypoglycaemia. Br Med J, 296, 949–50. Schade DS (1993) Successful treatment of severe 310 CH 21 HYPOGLYCAEMIA

refractory sulfonylurea-induced hypoglycemia with 114. Pegg A, Fitzgerald F, Wise D, Singh BM and octreotide. Diabetes Care, 16, 184–6. Wise PH (1991) A community-based study of 112. Boyle PJ, Justice K, Krentz AJ, Nagy RJ and Schade diabetes-related skills and knowledge in elderly DS (1993) Octreotide reverses hyperinsulinemia and people with insulin-requiring diabetes. Diabet Med, prevents hypoglycemia induced by sulfonylurea over- 8, 778–81. doses. J Clin Endocrinol Metab, 76, 752–6. 115. Strachan MWJ, Deary IJ, Ewing FME and Frier BM 113. Mutch WJ and Dingwall-Fordyce I (1985) Is it a (1997) Is type II diabetes associated with an increased hypo? Knowledge of symptoms of hypoglycaemia in risk of cognitive dysfunction? A critical review of elderly diabetic patients. Diabet Med, 2, 54–6. published studies. Diabetes Care, 20, 438–45. 22 Diabetes in Care Homes

Alan Sinclair1 and Terry Aspray2 1Bedfordshire and Hertfordshire Postgraduate Medical School, University of Bedfordshire, Putteridge Bury, Hitchin Road, Luton, UK 2Newcastle University, Campus for Ageing and Vitality, Newcastle upon Tyne, UK

In the industrialized nations of the world, increasing Key messages numbers of older people continue to leave their own homes and move into institutions where their physi- • The quality of diabetes care within care homes cal and social needs may be better met. In the United and other assisted-living facilities needs to States, apart from chronic care settings such as nursing improve and this requires implementation of homes, there is a growing industry of assisted-living fa- evidence-based guidelines and policy enforce- cilities where facilities such as the provision of a meal ment service, supervision, leisure and cleaning are provided. • Screening for diabetes at the time of admission In other countries such as Australia, low-level hostels to a care home and regularly afterwards is an important measurable outcome provide care services for many older patients. The or- • Maintenance of health status and functional ganization and operation of such facilities affect the performance are key goals of care for all nature of the provision of health and social care to res- residents with diabetes idents and as a consequence impact on the quality of care delivered, irrespective of the health condition.

22.1 Introduction 22.2 The United Kingdom as a model of care home reform For the United Kingdom, between the years 2000 and 2050, it is projected that the number of people In the United Kingdom, after World War II, the 1948 aged over 65 years will continue to increase from 9.3 National Assistance Act established the local authority million to 17 million. In addition, a fourfold increase as the responsible body for overseeing the reform is anticipated in the number of over- 85-year-olds, of public assistance institutions (the origins of which from 1.1 million to 4.4 million! [1]. Independent living came from the Elizabethan Poor Law) and the creation may be increasingly difﬁcult for this older population of residential care homes for older people “...in need as frailty and chronic ill-heath are accompanied by of care and attention”. isolation. Diabetes is known to be an independent risk Improvements to bed provision were slow and ham- factor for admission to a care home [2] and is impicated pered by a lack of funds. Residents were more frail in up to a quarter of admissions [3]. and increasingly dependent, but fewer hospital beds

Diabetes in Old Age Third Edition Edited by Alan J. Sinclair © 2009 John Wiley & Sons Ltd. ISBN: 978-0-470-06562-4 312 CH 22 DIABETES IN CARE HOMES for frail older people were being provided, putting registered nurses who can provide care for more continued pressure on local authority and social ser- complex health needs. vices to support the care of frail older people at This distinction between social and nursing care needs home and in residential and nursing homes. During applies to individuals. Although residents may enter a the 1980s, there was also an expansion of the indepen- home requiring only personal care assistance, they are dent sector to complement (or compete with) social likely to become increasingly frail, and health needs services provision. Between 1982 and 1991, the num- may increase requiring a nursing bed. While this may ber of beds in private care homes rose from 49 900 to necessitate a move, many homes offer both residential 161 200 [4]. However, the free provision of care within and nursing care, allowing the resident to remain in the these homes is dependent on residents having a low same home. level of wealth. As increasing numbers of frail older Other aspects of specialist care include the accom- people have moved into private care homes, National modation of residents with mental health disorder. Health Service (NHS) long-stay hospital beds have also The archaic term Elderly Mentally Inﬁrm is abbrevi- closed, so that more than 50% of all health care beds ated to EMI, which remains in common usage. For in the UK are now provided in nursing homes! [5]. this group, both residential and nursing care is pro- As a result of these trends, more people are living vided. Finally, there is an increasing population of in care homes and, indeed, UK estimates are that older adults with learning disability, who have left their the current population of 450 000 will increase to former long-term care in psychiatric hospitals to live 1 130 000 in the next 50 years. This will be associated in the community and present their own special chal- with the social and health costs of providing care lenges [7]. escalating from £13 billion to £55 billion by the year 2051 [6]. In the UK, Section 49 of the Health and Social Services Act, 2001, made care provided by 22.3 Epidemiology registered nurses in care homes an NHS responsibility, and thus free of charge. Hence, the nomenclature is A number of studies and surveys of diabetes prevalence intimately linked with costs of care and eligibility for have been performed, and the key references for this free care funded by the NHS. are presented in Table 22.1. In the US, the proportion Since 2004, the Commission for Social Care In- of residents with diagnosed diabetes has increased from spection (CSCI) has incorporated the roles of the So- 14.5% to 24.6% between 1979 and 2004 [8]. This may cial Services Inspectorate (SSI), the joint review team represent an increase in prevalence of type 2 diabetes of the SSI, the National Care Standards Commission in the USA and possibly an increase in survival, but (NCSC) and the Audit Commission in the UK. Further there has probably been an increase in screening and reorganization of regulatory services is anticipated, in- diagnosis of the condition to explain much of this trend. corporating the Healthcare Commission and the Mental European estimates of self-reported diabetes are much Health act commission. However, the CSCI and its lower; however, when studies have been augmented successor will have responsibility for the regulation, by the direct assessment of glucose tolerance, the inspection and review of social care services. In partic- estimated prevalence exceeds 20%. Such studies thus ular, they inspect care homes and rate them, according identify a large population who are unaware of their to national standards, as well as reviewing local coun- abnormal glucose tolerance. cils and their own provision of advice and purchase of There are important challenges highlighted by these services. According to the UK Commission for Social data. If interested in diabetes prevalence, then we can Care Inspection, the distinction between Residential be convinced that the condition is common and – at and Nursing homes is that: least in Europe – often undiagnosed. If we wish to identify cases for treatment, then widespread screening • Residential care (sometimes termed personal care programmes for diabetes may not be practical, where only) is provided to people either short or long term, such a large proportion are frail and unlikely to comply offering accommodation, meals and personal care with a glucose tolerance test. The use of fasting glucose (such as help with washing and eating). alone will miss cases, and a postprandial test may be • Nursing care homes, by comparison, offer the same helpful in increasing the numbers screened but will as those without nursing care but they also have also miss cases. 22.4 COMPLICATIONS AND COMORBIDITY 313

Table 22.1 Estimates of diabetes prevalence among frail older care home residents. Source No. of patients Criteria Prevalence Comments (%) Wisconsin, USA, 7850 (approx.) Self reported 10 A community-based study: 0.9% of cases 1979 [9] living in nursing homes. However, 17.5% of adults aged 80+ years were in nursing homes! Toronto 1965–86 1177 Newly diagnosed 30–35 Longitudinal study of care home residents [10] not diabetic at admission on oral glucose tolerance test. US National Self-report 14.5 Difficulty in collating data from this report. Nursing Home survey, USA 1979 Wales, 1997 ([11]) 1514 Self-report 7.2 Study was not aimed at screening new cases but at looking at characteristics of known cases. Liverpool, 1997 1611 Self-report 9.9 Primarily a description of the quality of care. [12] Germany, 2001 1936 Self-report 26.2 Study also examined the undiagnosed [13] population, but used HbA1c as diagnostic test. Data not considered here. Birmingham UK 636 Self-report 12.0 The first study to use a screening test. WHO [3] Plus OGTT 26.7 1998 Criteria used. Details given of those unable to participate. USA, 2004 [8] 549 125 Self-report 26.4 This review of all new admissions to care homes during 2002 suggests that diagnostic coverage has improved. Newcastle, 2006 1461 Self-report 11.4 Fasting and postprandial glucose used to [14] Plus fasting/ 19.9 increase coverage but may have resulted in postprandial underestimate of prevalence. Difference in glucose prevalence between residential, nursing and EMI care homes found. New Zealand, 2006 1567 Self-report 11.7 [15] Norway, 2006 [16] 788 Self-report 20.2 Including frail older people in their own 186 HbA1c/OGTT 0.5 home BUT the majority (427) were unfit to participate. Patients had to have a raised HbA1c in order to be eligible for OGTT screening.

OGTT = oral glucose tolerance test. 22.4 Complications and than for those living at home. For example, fewer residents have been found to receive vaccination or comorbidity blood pressure monitoring [17]. In one large study of American nursing homes, diabetes, dementia, Chronic disease is common among care home cancer, heart failure, renal failure, chronic pulmonary residents, compared with older people living in the disease and anaemia were all associated with an community. Despite this, the quality of care for increased risk of mortality at 12 months [18]. older people living in nursing homes may be worse Susceptibility to infections, particularly pneumonia, 314 CH 22 DIABETES IN CARE HOMES has been highlighted in nursing homes [19–21] 22.5 Common management and, more recently, with concern about the risk of methicillin-resistant Staphylococcus aureus (MRSA) problems infection [22, 23]. Some common clinical management problems arise in Having diabetes is associated with a doubling of the the care of older adults, living in care homes (see risk of admission to nursing home from a subject’s Table 22.2): own home [2], and diabetes accounts for 12.3% of all admissions to nursing home care [24]. Disability • Nutrition: Weight loss and nutritional deficiency associated with diabetes is characteristically progres- can occur through anorexic symptoms and reduced sive in those requiring long-term care [25]. Residents calorific intake. Other contributing factors include with diabetes are likely to be at greater risk from mi- severe physical and cognitive impairment, as well crovascular disease [26] and visual impairment, with as neurological and gastroenterological disorders as- cataract prevalent in 80% and diabetic retinopathy in sociated with dysphagia, including stroke. In the 2.1% of all nursing home residents [27]. In the UK future, increasing numbers of residents are also an- and the USA, diabetic residents are younger and at ticipated with obesity and associated problems, thus increased risk of hospital readmission, cognitive im- exacerbating function and mobility, as discussed pairment, limb amputations and death, when compared above [34]. to non-diabetic residents [12, 28]. Pressure sores are more common in diabetic residents: in one prospective • Increased risk of hypoglycaemia: This condition may study of 14 607 residents, the odds ratio for developing occur in residents receiving sulphonylureas or in- a pressure ulcer, adjusted for other comorbidities, was sulin, through several predisposing factors. These 1.4 (95% CI: 1.2–1.8) [29]. include: (i) nutritional deficiency and weight loss; In addition, residents with diabetes have a range of (ii) cognitive impairment resulting in meals being other comorbidities. Although there are no studies of missed through poor memory and orientation; (iii) dementia specific to care homes, cognitive impairment anorexic conditions such as malignancy or infection; is associated with diabetes in older age [30, 31]. One and (iv) a lack of awareness of the symptoms and recent study of diabetes prevalence found the highest signs of hypoglycaemia by residents themselves or rates of undiagnosed diabetes in EMI residential care by care staff. The latter may be compounded by a homes [14], and the diabetes care for these residents in lack of monitoring of diabetes by residents and staff. particular failed to meet local and national standards. • The impact of comorbidities on diabetes care warrants Infections: Recurrent skin, chest and urinary further study. For example, cognitive impairment may infections may occur, especially if the control of result in patients being less able to monitor their blood glucose is not optimal. Infections themselves glucose levels or inject insulin; Parkinson’s disease has predispose the resident with diabetes to marked been associated with an increase in cost of diabetes hyperglycaemia or metabolic decompensation, care by up to 300%! [32]. and even to hyperosmolar non-ketotic coma or Obesity is a risk factor for diabetes and, in one US ketosis. study, the proportion of obese residents newly admitted to nursing homes rose from 15% to 25% over a 10-year Table 22.2 Management problems in period [33]. Obesity in middle age is associated with a care homes. 30% increased risk of admission to nursing homes af- • Nutritional deficiency and weight loss ter 25 years [34]. The associated risks of diabetes may • explain some of the secular trend in diabetes preva- Increased risk of hypoglycaemia lence in US nursing homes, which have already been • Infections discussed [35]. However, for nursing home residents • Urinary incontinence with diabetes, undernutrition is also commonly seen. • Pressure sores Half of all subjects in one study received lower dietary energy intakes than recommended [36], while in a • Leg and foot ulceration further study switching diabetic residents to the normal • Communication difficulties diet provided in the care home was not associated with • Increased risk of adverse drug reactions any significant deterioration in glycaemic control [37]. 22.6 ORGANIZATION OF DIABETES CARE IN RESIDENTIAL SETTINGS 315 • Urinary incontinence: This may be secondary to Table 22.2. For example, there is an increased like- hyperglycaemia, urinary infection, poor mobility or lihood of undernutrition, risk of hypoglycaemia and cognitive impairment. recurrent urinary infection. With immobility comes the risk of ulceration of the lower limbs and perineum, the • Pressure sores and leg or foot ulceration: These can healing of which may be impaired by the presence of lead to rapid deterioration and need for hospital diabetes. admission. Although, both the ADA and the NSF include • Communication difficulties: These can lead to un- sections about adults living in institutions, these pri- recognized diabetes care needs. Predisposing fac- marily relate to prisoners and not to care home res- tors include cognitive impairment, dysphasia and idents. In response to such guidelines and standards, dysarthria from cerebrovascular or other neurolog- health services are bound to develop policies to imple- ical disease, and sensory impairments such as visual ment change, and the danger of using such inappropri- and hearing loss. ate standards has been highlighted as vulnerable older people may suffer in the absence of targets specific to • Increased risk of adverse drug reactions: These can their needs [42]. However, they can also act as impor- occur because residents are often taking multiple tant audit tools, allowing the priorities of older people drugs for their diabetes and other coexisting diseases. to be highlighted. In one case study, a combination of Risks can be exacerbated by infrequent review of national, local and care home-specific standards was medication and lack of monitoring of renal and used to develop an audit tool to evaluate services for hepatic function. older people [43].

22.6 Organization of diabetes care 22.6.2 Developing standards for care homes in residential settings There are two main barriers to optimizing diabetes care 22.6.1 Setting standards for diabetes care in care homes. First, there are some clinical issues (as already outlined above; see Table 22.2) which are spe- The American Diabetes Association (ADA) reviews cific to older people, and may not be included in local its clinical practice recommendations, annually, and diabetes service design. Second, there are factors re- these are both comprehensive and up-to-date [38]. In lating to the care home itself; for example, there is Great Britain, the National Institute for Health and a well-recognized lack of sufficient staff training with Clinical Excellence (NICE) and the National Service few opportunities for continuing professional develop- Framework (NSF) for diabetes identify the key compo- ment in diabetes. There are high rates of staff turnover nents of diabetes care [39, 40], and similar guidelines in many homes, compounded by a large proportion of are available from other organizations and countries. unqualified staff with little experience or training to Overall, in the UK the introduction of standards has prepare them for looking after residents with diabetes. been credited with improvements in health care, espe- Finally, there is often a lack of available resources of cially when linked to payments to general practitioners staff time, catering services and equipment. In addi- through the Quality and Outcomes Framework (QOF) tion, there may be a lack of clear boundaries between [41]. The Diabetes NSF identified a series of standards medical and nursing responsibilities, which can be ex- to be applied to the care of children and adults with acerbated by poor communication. diabetes, and their relevance (or lack of relevance) Staff preparing meals and supervising residents at and/or implications for residents of care homes has meal time may lack a basic understanding of nutritional been addressed in Table 22.3. These may be directly principles. Communication difficulties between staff applicable to other health care organizations in differ- and residents may exist, which prevent needs being ent countries. met; these may be linguistic or cultural, or might Appropriate standards of care should take into con- reflect comorbidities, including neurological problems. sideration, where feasible, with specific concerns for Restrictive professional boundaries may prevent health diabetes in frail, older care-home residents. The com- care professionals from having specific inputs into mon clinical problems in this group, which may not care homes, especially within the independent sector. be seen in younger adults with diabetes, are listed in Quite clearly, the establishment of national standards 316 CH 22 DIABETES IN CARE HOMES

Table 22.3 UK National Service Framework Standards, 2001. Standards Summary Comments with regard to implementation for Care Home Residents Prevention of type 2 • Reduce risk of developing type Focus on intervention has been on reducing diabetes 2 diabetes prevalence of overweight and obesity in adults • Population as a whole and children. Lifestyle interventions are often not appropriate to older subjects, especially those • Reduce inequalities living in institutions.

Identification of people • Identify people who do not Strategy has been to increase awareness of with diabetes know they have diabetes symptoms, but these are less specific and relevant in frail older people. Follow-up, regular testing and opportunistic screening should be feasible. However, such services may not reach care home residents [17]. Empowering people • Encourage partnership in Partnership may be difficult, where residents have with diabetes decision-making, self- lost autonomy and may be cognitively impaired. management and maintain Structured education programmes (e.g. healthy lifestyle DESMOND XPERT) do not accommodate • Parents and carers should be people living in institutions. fully engaged

Clinical care of adults • High-quality care for patients There may be speciﬁc barriers to monitoring: with diabetes throughout lifetime residential homes without nursing staff may not • Optimize control of blood glu- monitor blood glucose, and there may be cose, blood pressure and other increasing dependence on district nurses. There risk factors may be difﬁculties in transporting frail residents to diabetes clinics in primary care and hospital settings. Clinical care of children • High-quality care and transi- There is an analogy to be drawn here with transition and young people tion between paediatric and between routine diabetes care and specialist care with diabetes adult services for older people living in institutions.

Management of diabetic • Protocols for treatment of di- No specific mention is made of the management of emergencies abetic emergencies and acute hyperosmolar states, more common in older complications people and associated with adverse outcomes. • Minimize risk of recurrence Care-home staff will require specific training on the identification and treatment of hypoglycaemia. Care of people with • Effective hospital care for dia- These apply equally to frail older people admitted diabetes during betes to hospital. admission to hospital • Continue to be involved in decisions

Diabetes and pregnancy • Optimize outcome of preg- Hopefully, not applicable! nancy for women

Detection and • Surveillance for long-term Retinopathy and podiatry surveillance may not management of complications reach care home residents. Less aggressive long-term • Protocols to investigate and to approaches to secondary prevention may be complications treat reducing risk of disability adopted, through inertia and/or lack of evidence. and premature death Long-term benefits may not be as relevant to frail older patients with a shorter life expectancy. 22.6 ORGANIZATION OF DIABETES CARE IN RESIDENTIAL SETTINGS 317 of diabetes care within care homes may be an important Difficulties with diabetes care in institutional settings initiative to promote care within these settings. have been shown elsewhere. For example, in the There is a clear lack of diabetes-related experience United States Funnel and Herman examined the poli- and knowledge among care home staff, and appropriate cies and practices in a group of 17 skilled nursing education and training is needed to improve diabetes homes in Michigan [45]. Although, the ADA and the care. However, difficulties in providing education and American Association for Diabetes Education first de- training include a lack of staff training budget in many veloped guidelines for diabetes care in skilled nursing homes, resulting in a reliance on free advice and infor- homes in 1981 [46], the authors carried out their re- mation. Many care staff are young and unskilled, with view using the 1995 version of less-specific criteria older members of staff often employed part time and derived from the ADA [38]. The homes studied were unqualified. Many homes have a high staff turnover generally large (mean number of beds 137) and the rate, with poor pay and conditions, which can lead to number of residents with diabetes per home ranged a low staff morale. Nursing staff work a rotating shift from 1 to 46 (mean 19). Almost all of the homes system which can lead to a poor continuity of care, and reviewed had some diabetes care protocols, plans or often precludes attendance at training events. In spite standing orders in place, although the standing or- of these difficulties, diabetes training and education is ders usually consisted of guidelines relating to nutri- provided to homes by many local diabetes care teams. tion or some aspects of nursing care. Guidelines of These are usually welcomed by care home managers, care relating to parameters of metabolic control, when and their success seems to relate to good local relation- to call a physician, or the surveillance of complica- ships being built up. It also requires the local diabetes tions, were least often present. In general, the care team to feel a responsibility for these homes and to be provided did not meet local or national standards of allowed by their managers to go in and help. diabetes care, but care practices were better when In the United Kingdom, individual home proprietors registered dieticians were involved in meal planning and trade associations can assist in improving diabetes and where written institutional policies were actually care. For example, the Independent Healthcare Associ- present. ation is the largest association in the independent sec- More recently, a European Working Party devoted tor, representing acute, psychiatric and long-term care a section of its evidence-based diabetes guidelines providers across the UK. By facilitating the promotion specifically to care homes [47]. Such guidelines offer and dissemination of best practice, research reports and an important competing viewpoint to more generic quality control systems within care homes, they are guidelines such as the NSF, outlined in Table 22.3. well placed to liaise with care home owners, managers Examples of evidenced-based recommendations pro- and staff to support education and training initiatives. vided in the European Guidelines are as follows: In response to such concerns, Diabetes UK has reviewed diabetes care in institutions and produced a • At the time of admission to a care home, each res- guideline for care homes. The Diabetes UK guideline ident requires to be screened for the presence of [44] synthesized available evidence to identify the diabetes. Level of evidence 2++; Grade of recom- clinical issues of particular relevance to this group: mendation B. • Lack of care plans and case management approaches • Each resident should have an annual screen for di- for residents with diabetes. abetes. Level of evidence 2+; Grade of recommendation C. • Inadequate nutritional guidance. • Each resident with diabetes should have an individ- • Lack of specialist health professional input. ualized diabetes care plan with the following mini- • Diabetes care not coordinated between primary and mum details: dietary plan, medication list, glycaemic secondary care services. targets, weight and nursing plan. Level of evidence 2+; Grade of recommendation C. • Inadequate review and poor metabolic control. • Each care home with diabetes residents should have • Lack of knowledge relating to diabetes among staff an agreed Diabetes Care Policy or Protocol which is working in care homes. regularly audited. Level of evidence 2++; Grade of • Lack of structured education and training for staff. recommendation B. 318 CH 22 DIABETES IN CARE HOMES • Optimal blood pressure and blood glucose regula- • Preventing unnecessary hospital admissions for dia- tion may help to maintain cognitive and physical betic foot problems. performance for each resident with diabetes. Level of evidence 2+; Grade of recommendation C (ex- Eye Care trapolated data). • Preserving vision, including screening for refractive error and cataract. 22.6.3 Improving care Screening for Neurovascular Complications If current guidelines and protocols fail to support ap- • Peripheral neuropathy and peripheral vascular dis- propriate diabetes care in institutional settings, then ease which predispose to foot infection and ulcera- what should be the aims of optimal care? Residents tion. with diabetes in care homes should receive a level of Management of Coexisting Disease in a Structured comprehensive diabetes care commensurate with their Way needs, and this should be on an equitable basis with those people with diabetes who do not live in an insti- • Including the diagnosis and treatment of depressive tutional setting. The two most important objectives are: illness, heart failure and hypertension. 1. To maintain the highest degree of quality of life Nutritional Assessment and well-being, without subjecting residents to unnecessary and inappropriate medical and therapeutic • Provide a well-balanced individualized healthy eat- interventions. ing 2. To provide sufficient support and opportunity to en- • Compatible with well-being able residents to manage their own diabetes condi- • Maintaining appropriate bodyweight. tion where this is a feasible and worthwhile option. However, there are several additional processes of care 22.6.4 Nutrition in diabetic residents which represent important goals to achieve for any Residents are likely to have several reasons for being resident with diabetes in a care home. nutritionally at risk. These include a lack of nutritional Metabolic Control knowledge and outdated ideas about diabetic diets that • Avoiding malaise and lethargy of hyperglycaemia are held by some staff. It is vital that up-to-date information about diabetes and healthy eating be provided • Minimizing the risk of hypoglycaemia to care home staff, especially those who have responsi- • Promoting the greatest level of physical and cogni- bility for menu planning, food purchasing and cooking. tive function. Local dietetic services will usually be a good source of help and advice in implementing healthy-eating poli- The European Guidelines [47] provide the following cies. They may often be able to help in staff training recommendation for glycaemic control: on the dietary aspects of diabetes care.

“For frail (dependent; multisystem disease; care home residency including those with dementia) patients 22.6.5 Responsibility of the physician where the hypoglycaemia risk is high and symptom All residents of care homes in the UK are registered control and avoidance of metabolic decompensation is with a general practitioner (GP), and diabetes care is paramount, the target HbA1c range should be >7.5 to assumed to be delivered by GPs for the majority of ≤.5%. Evidence level 3/4; Grade of recommendation D.” (In the revision of Guidelines to be published in patients. The increasing numbers of older people in early 2009, a target range of <7.0% will be stated.) care homes is having a significant impact on the workload of many GPs [48], and there is no recognition Foot Care or encouragement for GPs to provide specialist diabetes care in residential settings. Many visits to care • Preserving the integrity of the feet homes are ‘reactive’ in nature, taking place only when • Promoting mobility where possible a problem has been identified by the home staff. Care 22.6 ORGANIZATION OF DIABETES CARE IN RESIDENTIAL SETTINGS 319 home residents often have mobility problems, prevent- • The plan should include: (i) a series of metabolic tar- ing them from visiting the GP’s surgery for annual gets; and (ii) an individualized dietary and nutritional review, and few GPs provide a multidisciplinary an- plan. nual review service in the care home. Diabetologists • An annual review assessment should involve: (i) an often have little experience of managing frail older eye check; and (ii) a foot check. people in the community, and geriatricians are doing less in continuing and community care. Meanwhile, • Support and assistance in diabetes care from a named commissioning priorities for older people focus on the person who will be involved in metabolic monitoring management of long-term conditions by community with the resident. matrons and other non-medical staff, concentrating on • acute illness and the prevention of hospital admission. Inclusion of all residents with diabetes in the local The transfer of long-term care from hospitals to care diabetes registers at General Practice and/or district homes has not been accompanied by any significant level, as appropriate transfer of medical resources to the community. In con- In the locally variable arrangements that exist in the sequence, older people in care homes increasingly fall UK, these elements may be provided by a number of between primary, secondary and social care services, health care professionals. The diabetes NSF covers all and all too often their needs are forgotten [49]. of these, but the targeting of this community is not well Since some concerns about care home medicine in addressed, and there is the risk that QOF returns are the UK were highlighted over 10 years ago [50], there not submitted for some of the more challenging tasks, have been no significant national developments and no resulting in older people being excluded from care. clear model has evolved. At the time, a number of options was envisaged: visiting medical officers; ded- 22.6.7 Nursing care icated geriatric medical and psychiatric outreach services; integrated care by specialists, commissioned by Diabetes specialist nurses primary care or a more formalized model of shared care These nurses, who have had special training and edu- between hospital and GP. Some studies have been con- cation in diabetes, are known to be an invaluable link ducted in the UK on chronic disease management us- between primary and secondary diabetes care for older ing American-style Health Maintenance Organizations people [52], and can provide a high-quality service to (HMOs), but they have not gone as far as employing disadvantaged people with diabetes [53]. Increasingly, their own medical staff. Rather perversely, the Evercare diabetes specialist nurses are employed to work in the project used a model of care devised and implemented community and, within the time constraints of their in the USA for the care of nursing home residents, busy jobs, may become involved in diabetes education but applied it to frail older people living in their own and support for home care staff, assisting in the devel- homes, with no evidence of beneficial effect [51]. opment of the diabetes care policies for the home and Today, the GPs are still responsible for the medical individual care plans. care of individual residents registered with their practice. There remains no formal structure for the routine involvement of consultants in geriatric medicine or di- Primary care practice nurses abetes, nor other health care professionals to provide Increasingly, practice-based nurses who have had spe- multidisciplinary diabetes care when required. In the cial training in diabetes are coordinating diabetes care absence of any formal national structure local, ad hoc in general practice. They may also be empowered to arrangements are still being employed in an attempt to visit residents of the practice who are living in care provide the best possible multidisciplinary care. homes, to assist in the delivery of the care objectives outlined above. 22.6.6 Multidisciplinary diabetes care The elements of multidisciplinary diabetes care include District (community) nurses the following: District nurses can play an immense supporting role • An individualized diabetes care plan, with each res- in diabetes care in residential settings, despite many ident contributing to agreed objectives summarized receiving little – if any – special training in this area. in a care plan. The major remit of the district nurse is in the provision 320 CH 22 DIABETES IN CARE HOMES of nursing support to residential homes, including ad- of detecting early signs of foot ulceration and infec- vice to staff on diabetes care. They often adminis- tion, so that urgent prompt referral and action can be ter insulin to residents unable to self-inject because taken. The local state-registered podiatrist with an in- of physical impairment, cognitive disability or be- terest in diabetes will usually be the best person to havioural disturbance. Specific arrangements can be provide this help. made between care homes and District Nursing services with the delegation of specific diabetes care tasks 22.6.9 Provision of eye care to care home staff. However, care homes have been The lack of specialist eye care and regular ophthal- criticised over recent years for their medicines manage- mology review of residents with diabetes has been ment policies, with insulin having attracted particularly demonstrated in UK care homes [12, 56]. Many older heavy criticism [54]. There can also be tensions, where people with diabetes have undetected refractive error, the residents’ personal care needs alter to a point where and screening of immobile residents in care homes is they require nursing rather than personal care, with an feasible, but costly [58]. expectation that the care home nursing staff will offer The national standard for eye screening programmes monitoring and insulin therapy. in the UK are established, including exclusion criteria [59]. Where screening programmes are based 22.6.8 Footcare on examinations carried out by experienced and specially trained optometrists, there may be a better ser- Published information from many countries worldwide vice to older people, as refractive error, glaucoma and testifies to the high prevalence of diabetic foot disease cataract can be checked at the same time as screen- among the residents of care homes [26, 55–57]. The ing for diabetic retinopathy. The national standard is risk of foot ulceration is increased in those with ad- for diabetes eye screening using digital photography of vancing age, loss of protective pain sensation due to the retina. However, immobile patients are excluded, diabetic peripheral neuropathy, peripheral vascular dis- which argues that they are unlikely to receive retinal ease, and bony foot abnormalities. Although residents surgery. should have access to free care from state-registered The barriers to optometrists working in care homes podiatrists, in some homes private podiatrists are em- include: ployed to offer routine foot care, and residents maybe • encouraged to pay fees for their footcare. Thus, a local The funding of retinal screening at the exclusion of state-registered podiatrist with an interest in diabetes is eye examinations in care homes. a very important member of a local multidisciplinary • No financially viable option for the self-employed diabetes team, and his or her skills need to be utilized optometrists. by care home staff in appropriate ways. All people with diabetes should have an annual foot Thus, eye care for care home residents could be im- examination as part of the review process, and residents proved by an adequate funding of optometric assess- in care homes are not exempt from this recommenda- ment by contractual arrangements with the local com- tion [39]. This examination is to detect feet at risk of missioners resulting in: ulceration. At its simplest, this involves a brief his- • An improved and regular access of optometrists into tory to discover any previous episodes of ulceration, care homes an inspection of the feet to check for bony abnormali- • ties, palpation of the dorsalis pedis and posterior tibial Visual screening of all new admissions who have diabetes. pulses to detect ischaemia, and the use of a 5.07 g nylon monofilament to detect the loss of protective pain This would require: (i) improved accommodation and sensation. This foot examination can be carried out facilities at each care home to allow full optometric by any member of the community diabetes team who assessment; (ii) the education of care staff about the has the relevant skills and experience and, if the foot visual health in residents; (iii) the identification of a is deemed to be at risk, it should be checked every member of care home staff responsible for organizing 3 months by a podiatrist. It is also important to train visits by the optometrist; and (iv) improved referral care home staff to understand the importance of pre- systems for residents with eye problems to specialist ventive footcare, and to alert them to the importance secondary care. 22.7 SUSTAINING EFFECTIVE DIABETES CARE 321

22.6.10 Assessing the efficacy and Table 22.4 Outcome measures for use in residential efficiency of diabetes care diabetes care. Outcome measurements for diabetes in Primary Care 1. The percentage of residents achieving agreed metabolic targets of HBA , blood pressure, and weight during have been incorporated into the audit tools supporting 1c previous 12 months the NSF [39], and some of the gaps relating to the care of older people are outlined in Table 22.3. A uni- 2. Frequency and severity of hypoglycaemic episodes in previous 12 months form, comprehensive, standardized assessment for the routine long-term care of older people, the minimum 3. Frequency of hospital admissions for diabetes-related dataset–resident assessment instrument (MDS-RAI), problems in previous 12 months has been introduced into all nursing homes in the 4. Complication rates of visual loss, foot ulceration, renal United States and Iceland, and also in three provinces impairment, and angina in Canada. A US research group has combined data 5. Changes in level of dependency and physical and mental from the MDS-RAI instrument with other available function using the Barthel ADL (or extended ADL data from Medicare and hospital discharge, to study measures) and Mini-Mental State Examination Score treatment effects using valid measures of outcome in (MMSE) during previous 12 months this frail population [60]. More recently, similar as- 6. Health-related quality of life and well-being of each sessments have been carried out in the UK to assess resident with diabetes (e.g. using the SF 36 or sickness nursing care needs [61]. impact profile (SIP) measures); changes from admission Today, a number of national and international out- to now, or changes within previous 12 months come measures are available for older adults with dia- 7. The percentage of patients with completed diabetes care betes [47, 56], but these have not (yet) been adequately plans and annual reviews in the past 12 months tested in care home settings. The purpose of outcome measures in care homes is to: 22.7.1 What the care home needs to • Assess the quality of care delivered to each resident provide with diabetes. In order to sustain effective diabetes care, homes need to provide a suitable care environment in terms of • Assess the impact of diabetes on each resident in staff, resources, equipment and facilities. These should terms of personal well-being, functional disability, include: and rate of diabetes complications. • Staff who have received appropriate training and • Determine the impact of use of care home resources education in the basic management diabetes in care for residents with diabetes in terms of use of care home settings. staff time, dietary planning, monitoring equipment, and educational initiatives. • Facilities to carry out blood glucose monitoring and staff trained in the use of the equipment. The potential outcome measures are summarized in • Accommodation for annual review examination of Table 22.4. The data collection must be carried out by residents and foot care. care staff and visiting health care professionals, and • A member of catering staff familiar with dietary must represent the common objectives of diabetes care planning for residents with diabetes. for all parties. • A protocol of diabetes care agreed by the staff of the home, visiting diabetes health care professionals 22.7 Sustaining effective diabetes and the GP. care • A method of collecting agreed diabetes outcome indicator data. Detailed recommendations for care are contained in both the BDA and the European Working Party reports • Sufficient staff members trained to administer in- [44, 47]. These are outlined below. sulin. 322 CH 22 DIABETES IN CARE HOMES • Educational resources on diabetes for residents and 22.8 Conclusions their families. • Access to transport to enable residents to receive Previously, diabetes care in residential settings has not specialist treatment off site. attracted a great deal of scientific clinical enquiry, and consequently little has been known about the quality of • An admission policy including a strategy for diabetes care delivered; or the outcomes of care in res- those with known diabetes and screening for idential settings. There are many important topics for diabetes. future clinical research as the number of people with diabetes resident in care homes across the world con- Aspray and colleagues have shown that the use of tinues to grow rapidly. The tremendous morbidity and a combination of such standards (generalized and disability of residents with diabetes within long-term age-environment-specific) can point to appropriate ser- care poses many complex and challenging problems vice changes [43]. for all health care professionals involved in delivering diabetes care. Some possible practical strategies 22.7.2 What needs to be provided at the to improve diabetes care have been proposed in this local level chapter. Local diabetes services must encompass the special needs of care home residents with diabetes, including References support and guidance for homes. In the UK, the general funding of care in residential and nursing homes will 1. Wittenberg, R., Comas-Herrera, A., Pickard, L. and remain a subject of continued government debate for Hancock, R. (2004) Future demand for long-term care the foreseeable future. The joint commissioning of in the UK: A summary of projections of long-term care health and social services for older people would be finance for older people to 2051. The Joseph Rowntree a great step forward. Diabetes contracts within this Foundation, York. context would support high-quality care for residents 2. Tsuji, I., Whalen, S. and Finucane, T. E. (1995). Pre- of care homes, and should include: dictors of nursing home placement in community-based long-term care. J Am Geriatr Soc, 43 (7), 761–6. • Optometric services providing both on-site and 3. Sinclair, A.J., Gadsby, R., Penfold, S., Croxson, S.C. clinic-based eye services. and Bayer, A.J. (2001) Prevalence of diabetes in care home residents. Diabetes Care, 24 (6), 1066–8. • Podiatry services with time specifically dedicated to 4. Means, R., Morbey, H. and Smith, R. (2002) From care home residents. community care to market care?: the development of welfare services for older people. Policy Press, Bristol. • Agreed criteria for referral to secondary and inter- 5. Pollock, A.M., Player, S. and Godden, S. (2001) How mediate care specialist services. private finance is moving primary care into corporate • At least one diabetes specialist nurse specifically ownership. Br Med J, 322 (7292), 960–3. responsible for older people. He or she would play 6. Paying for long-term care: Moving forward. (2006) a prominent role in the effective organization and The Joseph Rowntree Foundation, York. Available at: www.jrf.org.uk. delivery of diabetes care to care homes in the area, 7. Aspray, T.J., Francis, R.M., Tyrer, S.P. and Quilliam, including diabetes education. S.J. (1999) Patients with learning disability in the • At least one community dietician in each locality, community. Be Med J, 318 (7182), 476–7. responsible for dietary and nutritional support of 8. Travis, S.S., Buchanan, R.J., Wang, S. and Kim, M. residents. (2004) Analyses of nursing home residents with diabetes at admission. J Am Med Dir Assoc, 5 (5), 320–7. • The registration of all care home residents with 9. Klein, R., Klein, B.E., Moss, S.E., DeMets, D.L., Kauf- diabetes on diabetes registers to ensure that they are man, I. and Voss, P.S. (1984) Prevalence of diabetes involved in diabetes clinical audit projects. mellitus in southern Wisconsin. Am J Epidemiol, 119 (1), 54–61. • Diabetes educational and training programmes for 10. Grobin, W. (1989) A longitudinal study of impaired care home staff at local, regional and national level glucose tolerance and diabetes mellitus in the aged. J to ensure that the staff are kept up to date. Am Geriatr Soc, 37 (12), 1127–34. 22.8 CONCLUSIONS 323

11. Sinclair, A.J., Allard, I. and Bayer, A. (1997) Observa- 24. Russell, L.B., Valiyeva, E., Roman, S.H., Pogach, tions of diabetes care in long-term institutional settings L.M., Suh, D.C. and Safford, M.M. (2005) Hospitaliza- with measures of cognitive function and dependency. tions, nursing home admissions, and deaths attributable Diabetes Care, 20 (5), 778–84. to diabetes. Diabetes Care, 28 (7), 1611–17. 12. Benbow, S.J., Walsh, A. and Gill, G.V. (1997) Diabetes 25. Ferrucci, L., Guralnik, J.M., Pahor, M., Corti, M.C. in institutionalised elderly people: a forgotten popula- and Havlik, R.J. (1997) Hospital diagnoses, Medicare tion? Br Med J, 314 (7098), 1868–9. charges, and nursing home admissions in the year when 13. Hauner, H., Kurnaz, A.A., Haastert, B., Groschopp, C. older persons become severely disabled. JAMA, 277 and Feldhoff, K.H. (2001) Undiagnosed diabetes melli- (9), 728–34. tus and metabolic control assessed by HbA(1c) among 26. Mooradian, A.D., Osterweil, D., Petrasek, D. and Mor- residents of nursing homes. Exp Clin Endocrinol Dia- ley, J.E. (1988) Diabetes mellitus in elderly nursing betes, 109 (6), 326–9. home patients. A survey of clinical characteristics and 14. Aspray, T.J., Nesbit, K., Cassidy, T.P., Farrow, E. management. J Am Geriatr Soc, 36 (5), 391–6. and Hawthorne, G. (2006) Diabetes in British nursing 27. Whitmore, W.G. (1989) Eye disease in a geriatric nurs- and residential homes: a pragmatic screening study. ing home population. Ophthalmology, 96 (3), 393–8. Diabetes Care, 29 (3), 707–8. 28. Duffy, R.E., Mattson, B.J. and Zack, M. (2005) Co- 15. Gill, E.A., Corwin, P.A., Mangin, D.A. and Suther- morbidities among Ohio’s nursing home residents with land, M.G. (2006) Diabetes care in rest homes in diabetes. J Am Med Dir Assoc, 6 (6), 383–9. Christchurch, New Zealand. Diabet Med, 23 (11), 29. Berlowitz, D.R., Brandeis, G.H., Morris, J.N., Ash, 1252–6. A.S., Anderson, J.J., Kader, B., et al. (2001) Deriving a 16. Jorde, R. and Hagen, T. (2006) Screening for diabetes risk-adjustment model for pressure ulcer development using HbA1c in elderly subjects. Acta Diabetol,43(2), using the Minimum Data Set. J Am Geriatr Soc,49(7), 52–6. 866–71. 17. Fahey, T., Montgomery, A.A., Barnes, J. and 30. Korf, E.S., White, L.R., Scheltens, P. and Launer, Protheroe, J. (2003) Quality of care for elderly L.J. (2006) Brain aging in very old men with type residents in nursing homes and elderly people living 2 diabetes: the Honolulu-Asia Aging Study. Diabetes at home: controlled observational study. Br Med J, Care, 29 (10), 2268–74. 326 (7389), 580. 31. Xu, W.L., Qiu, C.X., Wahlin, A., Winblad, B. and 18. van Dijk, P.T., Mehr, D.R., Ooms, M.E., Madsen, R., Fratiglioni, L. (2004) Diabetes mellitus and risk of de- Petroski, G., Frijters, D.H., et al. (2005) Comorbidity mentia in the Kungsholmen project: a 6-year follow-up and 1-year mortality risks in nursing home residents. J study. Neurology, 63 (7), 1181–6. Am Geriatr Soc, 53 (4), 660–5. 32. Pressley, J.C., Louis, E.D., Tang, M.X., Cote, L., Co- 19. Garibaldi, R.A. (1999). Residential care and the el- hen, P.D., Glied, S., et al. (2003) The impact of comor- derly: the burden of infection. J Hosp Infect,43 bid disease and injuries on resource use and expendi- (Suppl.), S9–18. tures in parkinsonism. Neurology, 60 (1), 87–93. 20. Loeb, M. (2005) Epidemiology of community- and 33. Lapane, K. L. and Resnik, L. (2005) Obesity in nursing nursing home-acquired pneumonia in older adults. Ex- homes: an escalating problem. J Am Geriatr Soc,53(8), pert Rev Anti Infect Ther, 3 (2), 263–70. 1386–91. 21. Marrie, T.J., Durant, H. and Kwan, C. (1986) Nursing 34. Elkins, J.S., Whitmer, R.A., Sidney, S., Sorel, M., home-acquired pneumonia. A case-control study. JAm Yaffe, K. and Johnston, S.C. (2006) Midlife obesity Geriatr Soc, 34 (10), 697–702. and long-term risk of nursing home admission. Obesity 22. Libert, M., Elkholti, M., Massaut, J., Karmali, R., (Silver Spring), 14 (8), 1472–8. Mascart, G. and Cheriﬁ, S. (2008) Risk factors for 35. Resnick, H.E., Heineman, J., Stone, R. and Shorr, R.I. meticillin resistance and outcome of Staphylococcus (2008) Diabetes in U.S. nursing homes, 2004. Diabetes aureus bloodstream infection in a Belgian university Care, 31 (2), 287–8. hospital. J Hosp Infect, 68(1), 17–24. 36. Benbow, S.J., Hoyte, R. and Gill, G.V. (2001) Institu- 23. Lucet, J.C., Grenet, K., Armand-Lefevre, L., Harnal, tional dietary provision for diabetic patients. QJMed, M., Bouvet, E., Regnier, B., et al. (2005) High preva- 94 (1), 27–30. lence of carriage of methicillin-resistant Staphylococ- 37. Coulston, A.M., Mandelbaum, D. and Reaven, G.M. cus aureus at hospital admission in elderly patients: (1990) Dietary management of nursing home residents implications for infection control strategies. Infect Con- with non-insulin-dependent diabetes mellitus. Am J trol Hosp Epidemiol, 26 (2), 121–6. Clin Nutr, 51 (1), 67–71. 324 CH 22 DIABETES IN CARE HOMES

38. American Diabetes Association. (2008) Clinical Prac- 51. Gravelle, H., Dusheiko, M., Sheaff, R., Sargent, P., tice Recommendations. Diabetes Care, 31(Suppl. 1), Boaden, R., Pickard, S., et al. (2007) Impact of case S1–108. management (Evercare) on frail elderly patients: con- 39. Great Britain. Department of Health. (2001) National trolled before and after analysis of quantitative outcome service framework for diabetes: standards. Department data. Br Med J, 334 (7583), 31. of Health, London. 52. Sinclair, A.J., Turnbull, C.J. and Croxson, S.C. (1996) 40. National Institute for Clinical Excellence. (2002) Man- Document of care for older people with diabetes. Spe- agement of type 2 diabetes. National Institute for Clin- cial Interest Group in Diabetes, British Geriatrics So- ical Excellence, London. ciety. Postgrad Med J, 72 (848), 334–8. 41. Campbell, S., Reeves, D., Kontopantelis, E., Middle- 53. Norman, A., French, M., Hyam, V. and Hicks, D. ton, E., Sibbald, B. and Roland, M. (2007) Quality of (1998) Development and audit of a home clinic service. primary care in England with the introduction of pay Journal of Diabetic Nursing, 2 (2), 51–4. for performance. N Engl J Med, 357 (2), 181–90. 54. Great Britain. Commission for Social Care Inspection. 42. Holt, R.M., Schwartz, F.L. and Shubrook, J.H. (2007) (2006) Handled with care? Managing medicines for Diabetes care in extended-care facilities: appropriate residents of care homes and children’s homes Ð a follow intensity of care? Diabetes Care, 30 (6), 1454–8. up study. CSCI, London. 43. Aspray, T.J., Nesbit, K., Cassidy, T.P. and Hawthorne, 55. Cantelon, J.F. (1972) Diabetic residents of homes for G. (2006) Rapid assessment methods used for the aged: observations for an eleven-year period. JAm health-equity audit: diabetes mellitus among frail Geriatr Soc, 20 (1), 17–21. British care-home residents. Public Health, 120 (11), 56. Sinclair, A.J., Turnbull, C.J. and Croxson, S.C. (1997) 1042–51. Document of diabetes care for residential and nursing 44. British Diabetes Association. (1999) Guidelines of homes. Postgrad Med J, 73 (864), 611–12. Practice for Residents with Diabetes in Care Homes. 57. Wolffenbuttel, B.H., van Vliet, S., Knols, A.J., Slits, British Diabetes Association, London. W.L., Sels, J.P. and Nieuwenhuijzen Kruseman, A.C. 45. Funnell, M.M. and Herman, W.H. (1995) Diabetes (1991) Clinical characteristics and management of dia- care policies and practices in Michigan nursing homes, betic patients residing in a nursing home. Diabetes Res 1991. Diabetes Care, 18 (6), 862–6. Clin Pract, 13 (3), 199–206. 46. Van Nostrand, J. (1985) Nursing home care for di- 58. Anderson, S., Broadbent, D.M., Swain, J.Y., Vora, J.P. abetes. In: Diabetes in America: Diabetes data com- and Harding, S.P. (2003) Ambulatory photographic piled by National Diabetes Data group. US Govern- screening for diabetic retinopathy in nursing homes. ment Printing Ofﬁce, Washington, DC, NIH publication Eye, 17 (6), 711–16. 85-1468. 59. UK National Screening Committee. (2006) Exclud- 47. Sinclair, A.J., Cromme, P.V.M., Rodriguez-Manas, ing patients from the NHS Diabetic Retinopathy L., Fasching, P., Muggeo, M. and Hader, C. (2004) Screening Programme temporarily or perma- Clinical Guidelines for Type 2 Diabetes Mellitus nently: Good practice guide version 2.0, from in older people. Retrieved 1st July, 2008, from http://www.nscretinopathy.org.uk/exclusions.html http://www.eugms.org/index.php?pid=31. 60. Carpenter, G.I., Bernabei, R., Hirdes, J.P., Mor, V. and 48. Kavanagh, S. and Knapp, M. (1998) The impact on Steel, K. (2000) Building evidence on chronic disease general practitioners of the changing balance of care in old age. Standardised assessments and databases for elderly people living in institutions. Br Med J, 317 offer one way of building the evidence. Br Med J, 320 (7154), 322–7. (7234), 528–9. 49. Bowman, C., Johnson, M., Venables, D., Foote, C. 61. Carpenter, I., Perry, M., Challis, D. and Hope, K. and Kane, R.L. (1999) Geriatric care in the United (2003) Identiﬁcation of registered nursing care of res- Kingdom: aligning services to needs. Br Med J, 319 idents in English nursing homes using the Minimum (7217), 1119–22. Data Set Resident Assessment Instrument (MDS/RAI) 50. Black, D. and Bowman, C. (1997) Community insti- and Resource Utilisation Groups version III (RUG-III). tutional care for frail elderly people. Br Med J, 315 Age Ageing, 32 (3), 279–85. (7106), 441–2. 23 Primary and Community Care of Diabetes in Older People

Roger Gadsby Warwick Medical School, University of Warwick, Coventry, UK

Key messages 23.1.1 Definitions of primary, secondary and tertiary care • Today, with rising numbers of older people living • Primary care is defined as the first-contact, contin- with diabetes, there is a shift from secondary to primary care delivery in many countries. uous, comprehensive, coordinated care provided to • Multidisciplinary teamwork and pay-for- people undifferentiated by age, gender, disease or performance schemes have been developed to organ system [2] In most countries it is provided provide incentives for better care. by primary care physicians – often termed general • Care must be individualized, and issues of practitioners (GPs) – and their associated teams. polypharmacy and medication adherence ad- • Secondary care is hospital care, either as an inpatient dressed. or as an outpatient. Such care is often consulta- tive, short-term in nature, and provides assistance in the diagnosis and management for primary care 23.1 Introduction physicians. Secondary care is provided by specialist doctors (often termed consultants) and their special- The number of people worldwide who are living with ist teams. diabetes is rapidly increasing, with numbers expected to rise from today’s figure of 194 million – equal to • Tertiary care is very specialized care for those people 5.1% of the adult population – to a total of 333 million with specific conditions that are often complex and in 2025, or 6.3% of the adult population [1]. Since rare, and require specific expertise. Tertiary care is people in general are living longer, the number of usually based at regional or subregional hospitals, older people living with diabetes is likely to increase and provided by specialist doctors and their teams. significantly in most countries of the world. Inevitably, this will place a severe strain on healthcare resources 23.1.2 The funding and organization of and healthcare provision. Consequently, systems of healthcare care will need to evolve in order to cope with this increasing diabetes population, and in many countries Different countries have different models by which this has resulted in a shift in care status, from secondary health care is funded and organized. In the United to primary. Kingdom, care is free at the point of delivery and paid

Diabetes in Old Age Third Edition Edited by Alan J. Sinclair © 2009 John Wiley & Sons Ltd. ISBN: 978-0-470-06562-4 326 CH 23 PRIMARY AND COMMUNITY CARE OF DIABETES IN OLDER PEOPLE for by the government out of taxation. Primary care During the 1970s, some family doctors in the UK is delivered by GPs, usually working in partnership and in some other countries began to assume respon- with one another in small groups called ‘practices’. sibility for the routine review of their patients with The GPs employ their own staff and practice nurses. diabetes. This came about when the hospital doctors In other countries, however, funding is provided by developed a policy of trying to discharge patients to health insurance companies, and patients may pay via general practice in order to reduce the size of their a fee for service payment which is reimbursed by the clinics. Occasionally this occurred by agreement and insurance company. planning, but in other situations it just ‘happened’. Funding issues may inﬂuence the mode of provision Over the past 20–30 years, increasing numbers of of healthcare, the two central features of many primary people have been diagnosed with diabetes, and this led care-driven healthcare systems being the concepts of: to the hospital specialist model coming under considerable strain. In the UK, this was highlighted in a report • All people being registered with an individual pri- from the audit commission entitled ‘Testing Times’, mary care physician or practice which provides the which reviewed secondary care diabetes services and ﬁrst-contact care for all conditions. concluded that many services were under considerable strain. One way forward proposed in the report was • A ‘gatekeeper function’ which restricts access to more expensive and invasive secondary care services for primary care to provide more of the routine care, for those people referred from primary care [3]. so that hospitals could concentrate on specialist care and professional support and training, while allowing patients to receive continuity of care closer to home [4]. 23.2 The shift of diabetes care from Continuity of care with the same doctor and nurse the hospital to the community team is especially important for elderly people with diabetes, and this may be more easily provided in During the 1950s and 1960s, diabetes was seen in primary care (Figure 23.1). many countries to be a condition that was dealt with There is an evidence base to support this secondary by secondary care. Consequently, many hospitals had to primary care shift. In 1998, a study was conducted large diabetes outpatient clinics which people with which aimed to identify and evaluate all published ran- diabetes attended, often on an annual basis. domized trials of hospital versus general practice care

Hospital-based specialist Community Community Community dietician (diabetologist or podiatrists physiotherapists geriatrician with an interest) and team

GP and practice Older person Family and team with diabetes carers

Retinal screening services

Figure 23.1 Healthcare professionals involved in the care of older people with diabetes in the community. 23.4 ROUTINE CARE OF PEOPLE WITH DIABETES BY THE PRIMARY CARE DIABETES TEAM 327 for people with diabetes, so as to compare the effec- • Patients with unstable cardiovascular disease (CVD). tiveness of general practice and hospital care through • Patients with poor metabolic control where it is the use of meta-analysis of the identified trials [5]. Five proving very difficult to control HBA , lipids or trials that included a total of 1058 patients were iden- 1c blood pressure to agreed targets. tified that fulfilled the inclusion criteria; three of these were from the UK, and two from Australia. The overall mean age of the patients was 58.4 years, and the results were found to be heterogeneous between the 23.4 Routine care of people with trials. In shared care schemes featuring more intensive diabetes by the primary care support through a computerized prompting system for diabetes team GPs and patients, there was no difference in mortality between care in hospital and care in general practice. In a number of countries, the health economies The glycated haemoglobin (HbA1c) level tended to be have introduced – or are considering the introduction lower in primary care, and losses to follow up were of – financial incentives for providing good quality significantly lower in primary care. Schemes with less care for people with chronic diseases. Payments are well-developed support for family doctors were asso- often made when processes of care are shown to reach ciated with adverse outcomes for patients. a pre-specified standard. Some systems also reward The review concluded that unstructured care in the achievement of good quality care in specific areas. the community was associated with a poorer follow In diabetes, ‘pay for performance’ systems can reward up, worse glycaemic control and a greater mortality the achievement of pre-specified target levels of blood than hospital care. Computerized central recall, with pressure, HBA1c levels and lipids in a pre-specified prompting for patients and their family doctors, can percentage of patients. Although a number of pay for achieve standards of care as good as – or even bet- performance incentive schemes have been launched in ter than – hospital outpatient care, at least in the short the USA, the best-developed is in the UK, where it is term. The evidence supported the provision of a reg- referred to as a ‘Quality and Outcomes framework’. ular, prompted recall and the review of selected peo- In this system, points are achieved for both process ple with diabetes by willing GPs. Moreover, this goal and outcome achievement; the points then attract could be achieved if a suitable organization were in a payment figure which depends on the number of place. patients in the practice and the square root of the disease prevalence. When it was launched as part of the new GP con- 23.3 The primary care diabetes tract in the UK, the Department of Health stated team that the Quality and Outcomes Framework (QOF) represented – for the first time in any large health sys- This should consist of a motivated GP (primary care tem in any country – that GP practices would be re- physician) with access to a team of health care profes- warded systematically on the basis of the quality of sionals including a practice nurse, dietician, podiatrist care delivered to patients [7]. The basis for the Quality and a diabetes nurse or district (community) nurse with and Outcomes framework is a series of clinical do- well-defined responsibilities for each team member [6]. mains which contain 10 disease areas, for which there The practice should operate a dedicated record- are a maximum of 550 points. keeping system, with quality assessment and a recall For the first two years of the new GP contract in system to enhance clinic attendance. 2004/5 and 2005/6, the diabetes area had a maximum There should also be a structured protocol to include of 99 points (approximately 10% of the total available), criteria for referral to a hospital specialist. This should spread across 18 clinical indicators. include [6]: 23.4.1 Quality diabetes clinical indicators • Patients with severe vascular complications. relevant to older people • Patients who require treatment for diabetes eye dis- These indicators insist that: ease, foot ulceration or nephropathy. • a comprehensive diabetes review, including vision • Patients with increasing dependency and immobility. and footcare is undertaken annually; 328 CH 23 PRIMARY AND COMMUNITY CARE OF DIABETES IN OLDER PEOPLE • metabolic targets for blood pressure, HBA1c, and that would ‘outweigh’ any benefits gained in increasing cholesterol are individualized and agreed with the the quantity of life. older person, their families and carers; and This balance between striving for targets and providing the best care for an individual older person with • issues of polypharmacy and concordance with med- diabetes can be one of the most difficult judgements ications are discussed, and agreed. to make in the primary and community care of older people with diabetes. There has been an increase in the average number The quality and outcomes targets in the UK system of points scored in 2005/6 as compared to the previ- are based on evidence from randomized controlled ous year (2004/5), and also an increase in the register trials and guidelines; there is, however, a judgement to prevalence of diabetes, from 3.4% to 3.6%. There is ev- be made as to whether this evidence applies to older idence therefore that health care professionals – when people. provided with an incentive – will strive to achieve that incentive as they will achieve more income. A recently published study described changes in performance of 23.5.1 Primary prevention of adverse a random sample of 42 UK general practices in 1998, events: The evidence base 2003 and 2005 [8]. This reported that, although the Glycaemic control quality of care of people with diabetes was improving before the introduction of the ‘pay for performance’ In the United Kingdom Prospective Diabetes Study scheme in 2004, there was an acceleration in improve- of glycaemic control [10], newly diagnosed patients ment after its introduction. with type 2 diabetes were enrolled up to the age of In a review of six reports on the quality of diabetes 65 years and followed for mean of 10 years. The care in the UK conducted since 1999, there was a trend intensively treated group attained a mean HBA1c of 7% compared to the standard control group (mean towards improvement in both process and outcome 7.9%). This 0.9% reduction resulted in a 12% reduction of care relating to the period 1999–2004 [9]. The for any diabetes-related endpoint, and 25% reduction quality of both process and outcomes of care in the for microvascular endpoints. There was also a 16% Quality and Outcomes framework results for 2004/5 reduction in macrovascular events, which just failed and 2005/6 was significantly greater than that found to reach the level of statistical significance. Thus, it in the published studies from previous years, and might be suggested that there is an evidence base for there was improvement in the scores for all clinical tight glycaemic control up to the age of 75 years. indicators for diabetes from 2004/5 to 2005/6 (see Studies conducted specifically in older people tend to Table 23.1). The report authors concluded that modest be descriptive rather than true randomized controlled financial incentives paid to primary care physicians trials, and have shown an association between poor might represent a successful strategy for improving the glycaemic control and mortality [11], cardiovascular care of people with diabetes. events [12] and retinopathy [13].

23.5 Which strategies to prevent Cholesterol control CVD are appropriate for older In the Heart Protection Study (HPS) [14], patients aged people with diabetes? up to 80 years were recruited and followed for a mean of 4 years. Hence, it could be suggested that there is One of the main questions in such incentive schemes an evidence base for cholesterol lowering with a statin is how hard should health care professionals strive so up to 84 years. In the HPS, treatment with simvastatin that old and very old people achieve the quality targets (40 mg per day) resulted in a 27% reduction in the for glycaemic control, blood pressure and cholesterol. incidence of first non-fatal myocardial infarction, and The concern is that, by zealous striving after blood a 25% reduction in first incidence of fatal or non-fatal pressure, cholesterol and HBA1c targets, there will be stroke as compared to placebo. an impairment of the quality of life of older people with In the CARDS study [15], a total of 2838 people diabetes by increasing the risks of hypotension, the aged 40–75 years with type 2 diabetes and no coro- adverse side effects of medication and hypoglycaemia nary heart disease (CHD) but who had one other risk 23.5 WHICH STRATEGIES TO PREVENT CVD ARE APPROPRIATE FOR OLDER PEOPLE WITH DIABETES? 329 2005/06 2004/05 Difference 277,317 265,623 95.8 93.2 2.6 161,211 138,292 85.8 82.1 3.7 Denominator Numerator % % 1,651,515 1,477,561 89.5 85.2 4.3 1,870,940 1,821,376 97.4 95.91,841,5711,674,231 1,776,415 1.5 96.51,786,114 1,034,293 94.4 61.81,781,716 1,633,981 58.8 2.1 91.51,785,322 1,580,830 89.4 3.0 88.71,782,667 1,574,374 83.4 2.1 88.21,874,539 1,558,411 78.9 5.3 87.4 1,840,954 77.61,684,327 9.3 98.2 97.01,857,309 1,396,760 9.8 82.9 1,777,422 1.2 70.9 95.71,849,405 93.0 12.0 1,703,389 1,764,280 2.7 95.4 1,345,409 92.7 79.0 71.8 2.7 7.2 diabetes; clinical indicators for 2004/5 and 2005/6. lood pressure is 145/85 or less 17 55% 1,753,856 1,313,740 74.9 70.3 4.6 rd BMI in the previous 15 months 1,835,480 1,726,599 94.1 90.6 3.5 ver smoked where smoking status should be ave had influenza immunisation in the preceding 1 a record of HbA1c or equivalent in the previous 15 o have a record of serum creatinine testing in the whom the last HbA1C is 10 or less (or equivalent Quality and outcomes framework (QOF) scores for September to 31 March recorded once months previous 15 months except those who have ne smoking cessation advice has been offered in the last 15 months test/reference range depending on local laboratory) in last 15 months months in the previous 15 months months months with ACE inhibitors (or A2 antagonists) months months test/reference range depending on local laboratory) in last 15 months previous 15 months (exception reporting for patients withprevious proteinuria) 15 months Diabetes Mellitus (Diabetes) Quality indicator Diabetes Prevalence (registered) 53,211,253 1,890,663 3.6 3.4 0.2 Table 23.1 DM2DM3 The percentage of patients The with percentage diabetes of whose patients notes with reco diabetesDM4 in whom there is a record The of percentageDM5 smoking of status patients in with the diabetes who smoke The and percentageDM6 whose of notes diabetic contain patients a who record have that The percentageDM7 of patients with diabetes in whom The the percentageDM8 last of HbA1C patients is with 7.4 diabetes or in less The (or percentage equivalent DM9 of patients with diabetes who have The a percentageDM10 record of of patients retinal with screening diabetes in The with the percentage a previous of record 15 DM11 patients of with presence diabetes or with absence a of The record peripheral percentage of pulses of neuropathyDM12 patients testing with in diabetes the who previous have The 15 a percentage record of of patients the with blood diabetes pressure in inDM14 whom the the past last 15 b The percentage ofDM15 patients with diabetes wh The percentage ofDM16 patients with diabetes with proteinuria The or percentage micro-albuminuria of whoDM17 patients are with treated diabetes who have The a percentage record of ofDM18 patients total with cholesterol diabetes in whose the last The previous measured percentage 15 total of cholesterol patients within with previous diabetes 15 who h DM13 The percentage of patients with diabetes who have a record of micro-albuminuria testing in the 330 CH 23 PRIMARY AND COMMUNITY CARE OF DIABETES IN OLDER PEOPLE factor for CHD (e.g. hypertension or smoker) were ran- There is, therefore, a significant question as to the domized to atorvastatin (10 mg per day) or placebo. applicability of these trials to older people with type 2 The trial was stopped earlier than expected because the diabetes who consult in primary care. pre-specified early stopping rule for efficacy had been met; hence, the median duration of follow-up was 3.9 23.5.2 Applying the evidence to years. Compared with placebo, the risk reduction in independent and frail older people the atorvastatin group for a CHD event was 37%, and It is possible to divide old and very old people into for death was 27%. two broad categories: • Blood pressure-lowering Those who have type 2 diabetes as their only significant disease and are otherwise fit, healthy and The United Kingdom Prospective Diabetes Study living independently. Approximately one-third of in- (UKPDS) [16] began as a study of glycaemic control dividuals fall into this group, based on data obtained in people newly diagnosed with type 2 diabetes, from a large community study in Wales conducted but subsequently had a blood pressure control study during the 1990s, where the objective measures of embedded into it. People were recruited up to the age dependency were based on Barthel Activities of of 75 years and followed for a mean of 10 years. Daily Living (ADL) score, Extended ADL score In the blood pressure study, 1148 people with hyper- and Mini-Mental State Examination score [18]. The tension and type 2 diabetes were randomized to a tight recommendation for primary prevention in these sub- control arm or a less-tight control arm. The final mean jects, where the precise evidence is lacking, would be difference between the two groups was 10/5 mmHg to treat to targets set for younger people, in consulta- (144/82 mmHg with tight control, 154/87 mmHg with tion with the individual. Thus there would be striving less-tight). Over nine years, patients in the tight control to obtain all the QOF targets for that individual. group showed significant reductions in morbidity and • mortality, with: (i) a 32% reduction in diabetes-related Those who are frail and elderly and have significant death; (ii) a 44% reduction in fatal and non-fatal stroke; comorbidities, such as arthritis, high dependency lev- (iii) a 56% reduction in congestive cardiac failure; and els or significant dementia. Approximately two thirds (iv) a 37% reduction in developing microvascular com- of individuals fall into this group. The practical care plications. The tightly controlled patients were treated for people in this group would be to ensure symp- with the beta-blocker atenolol or the ACE-inhibitor tomatic control, avoid hypoglycaemia and intensive captopril, but the study was not sufficiently powered monitoring, in consultation with the individual and to determine which agent was superior [1]. their carers. In the Hypertension Optimal Treatment Trial [17], The GP contract QOF allows individuals to be ex- a total of 18 790 patients with hypertension were cluded for varying reasons, and exception reporting can randomized into three groups, the aim being to achieve therefore be used for older people if the clinician feels diastolic pressures less than 90 mmHg, 85 mmHg and that achieving targets is not in the individual patient’s 80 mmHg in each group. The trial included about best interests. 1500 people with type 2 diabetes, who were aged up to 80 years. There were significant reductions in 23.5.3 Exception reporting cardiovascular morbidity and mortality in the tightest Exception reporting allows the practice to exclude indi- controlled group, with a relative risk reduction of 50% vidual patients from the disease indicators in particular [2]. circumstances. These are: Based on the results of these studies, it can be suggested that there is an evidence base for blood • Patients excepted from the whole clinical area: pressure lowering up to the age of about 80 years in ◦ Those who have been recorded as refusing to people with type 2 diabetes. attend a review, having been invited on at least Unfortunately, the people enrolled into clinical trials three occasions during the preceding 12 months. tend to be relatively fit and healthy, with a single ◦ Those for whom it is not appropriate to review specific disease that is the ‘target’ of the trial. But the chronic disease parameters due to specific this does not correlate with many of the old and very circumstances, such as extreme frailty or terminal old people with type 2 diabetes seen in primary care. illness or severe dementia. 23.7 DELIVERING APPROPRIATE DIABETES CARE 331 ◦ Where a patient does not agree to investigation As none of the studies (apart from PROSPER) in- and treatment and, after a reasonable discussion or cluded elderly people, Mangin et al. concluded that written advice, has given their informed dissent; it the more likely reason for the results seen in el- is essential that such dissent has been recorded in derly people was a substitution of the cause of death. the medical notes. These authors argued that the best interests of elderly • Patients exempted from one clinical indicator only people might lie in investing money in health care (if a valid Read code is used): to relieve suffering, such as cataract operations, de- ◦ Patients on maximum tolerated doses of medica- mentia care and joint replacements, rather than the tion whose level of outcome remains suboptimal. primary prevention of CVD. Mangin et al. consid- ◦ Patients for whom prescribing a medication is ered this the reason why GPs in the UK during the not clinically appropriate; for example, those who first year of the Quality and Outcomes framework have an allergy, another contraindication, or who were reluctant to follow the guidelines for cholesterol have experienced an adverse reaction. measurement and lipid lowering in people aged >75 ◦ Where a patient has not tolerated a medication. years [21]. ◦ Where a patient does not agree to investigation In a large cohort study in Canada of people aged ≥66 and treatment and, after a reasonable discussion or years with newly diagnosed type 2 diabetes, 21.8% re- written advice, they have given their informed dis- ceived an antiplatelet drug, 39.6% an antihypertensive sent; this dissent must be recorded in the medical drug, and 21% a lipid-lowering drug within one year of notes. initiation on antidiabetic oral therapy [22]. Only 7.6% ◦ Where the patient has a supervening condition received all three agents. Whether this was a subop- which makes treatment of their condition inappro- timal use of cardioprotective drugs, or whether it was priate, for example cholesterol reduction where the appropriate prescribing clearly depended on the per- patient has liver disease. ceived risk:benefit ratio of these interventions in this ◦ Where an investigative service or secondary care population. service is unavailable. Mangin et al. argued that many patients fear the manner of their dying rather than death itself, and • Patients exempted automatically from any of the despite the distressing nature of some cardiac deaths indicators by reporting software: ◦ Patients newly diagnosed within the practice with many people regard death from a heart attack as a diabetes or who have recently registered with the “good way to go”. By using treatments for primary practice, who should have measurements made prevention, it is suggested that the GP is selecting for within 3 months and delivery of clinical standards another cause of death unknowingly, and doing this within 9 months; for example, blood pressure or without the patient’s informed consent. cholesterol measurements within target level. The argument continues that a way is needed to assess prevention and treatment of risk factors in the elderly that takes a wider perspective when balancing 23.6 Preventive health care in older benefits and harms. Rather than consider absolute risk people: Further debate and death prevention, consideration should be given to an overall life extension and a reduction in overall mor- The whole problem of preventive health care in older bidity, taking the duration of treatment into account. people has been recently highlighted in a report in the British Medical Journal [19]. In an analysis of the PROSPER trial, in which older people aged 70–82 23.7 Delivering appropriate years were randomized to pravastatin or placebo, there diabetes care was a clear but small benefit on mortality and morbidity from CVD in the group given pravastatin. How- Today, primary care organizations in most developed ever, all-cause mortality was not altered. The reduc- countries play a significant part in the chronic disease tion in cardiovascular-related deaths was countered by management of people with diabetes. Pay for perfor- an increase in deaths from cancer. A meta-analysis mance schemes provide an added incentive to deliver of all statin trials has failed to show any increase high-quality care. However, care delivery is usually in cancer deaths in people treated with statins [20]. structured around the patient with diabetes attending 332 CH 23 PRIMARY AND COMMUNITY CARE OF DIABETES IN OLDER PEOPLE the practice premises. For those people who are house- 23.7.1 Optimizing care and reducing bound, or who live in care homes, the unstructured, hospital admissions in the frail largely reactive care received results in poor levels of elderly the collection and processing of data, and reduced levels of diabetes care. In the author’s practice, a review Many health economies in the developed world have of non-attenders at the practice diabetes clinic showed realized that reducing the number of admissions to that a significant number were housebound or living in hospital of frail elderly patients with multiple chronic institutional settings [23]. conditions (which includes diabetes) is a vital step in In a recent re-audit of the 452 people on the practice trying to stem the rapid escalation in health care costs. diabetes register, a ‘housebound’ person was defined as One model is based on a system from the USA, which somebody for whom the last three GP contacts were uses nurses in a managed-care programme called ‘Ev- home visits, and who the practice would expect to visit ercare’. In one evaluation study directed specifically at home if such a request were made. The subgroup at long-stay nursing home residents [25], the provi- totalled 38 people, representing 8.2% of the practice sion of a case manager reduced hospital admissions by diabetes register. Of these patients, six lived in resi- 50% over a 15-month period compared to controls. The dential or nursing homes (1.3% of the practice diabetes Evercare programme is currently being trialled in the register), three lived in sheltered housing complexes UK, with a plan to recruit 3000 case managers, called and 29 lived in private housing. ‘community matrons’. The system combines elements Among the 38 housebound patients, two had ter- of proactive nurse-led assessment and intensive case minal illness and four had severe dementia (and were management, but located in the community rather than living in care homes), four were younger people with in a nursing home setting. chronic psychiatric illness and 26 were frail elderly in A recent study examined the rates of emergency ad- the age range 78–90 plus years. mission, emergency bed days and mortality between In order for those people who are housebound to 2001 and 2005 in nine Primary Care Trusts in the UK, receive optimal diabetes care, they either need special- who piloted the intervention [26]. A total of 62 Ever- ized support and transport to allow them to be taken care practices was compared with between 6960 and from their homes to the site of care delivery, or they 7695 control practices in England (depending on the analysis being carried out). The intervention showed need to receive diabetes care in their homes. no significant effect on rates of emergency admission, There is a trend towards a reduction in home visiting emergency bed days and mortality for a high-risk pop- among primary care teams in the UK (www.statistics ulation aged over 65 years with a history of two or 2002/3) which is being replicated in other developed more emergency medical admissions in the preceding countries. Most home visits now take place because a 13-month period. The authors concluded that the case request is made by the patient, their family or carer management of frail elderly people introduced an addi- because a health problem has developed. tional range of services into primary care, without any When such visits take place they often concentrate associated reduction in hospital admissions. on the presenting problem, and a review of any chronic conditions such as diabetes may be overlooked. Visits 23.7.2 Integration of diabetes care into the to patients in care homes are usually made in a similar way, in response to an acute problem. community One solution proposed in the UK in a review of In many health care systems the GP who has re- care home diabetes [24] was for the appointment of sponsibility for a defined number of patients is cen- a Diabetes Specialist Nurse, who would have respon- tral to the delivery of health care in the community sibility for older people with diabetes and who were (see Figure 23.1). In many health economies, GPs housebound or in a care home. The nurse would proac- now work alongside practice nurses, district nurses tively visit these people to provide diabetes care and and case managers for frail elderly patients. In car- review their diabetes management. The nurse would ing for older people with diabetes, community dieti- then be in a position to suggest appropriate changes cians, community diabetes specialist nurses, commu- in diabetes management to the GP. However, this nity physiotherapists, community podiatrists and those idea has not been taken up by health communities in working in retinal screening programmes are also vi- the UK. tal members of the care team. They also need to 23.8 PROBLEMS OF POLYPHARMACY AND MEDICATION CONCORDANCE IN OLDER PEOPLE WITH DIABETES 333 work alongside hospital-based specialists in geriatric • A drug ‘education’ – the reasons for taking each medicine, and their teams. Sharing information and therapy, the frequency of administration and a re- working together is vital if such a large, diverse group view of any side effects. of health care professionals is to function effectively so • Appropriate monitoring, for example renal function as to provide optimal care and reduce hospital admis- when prescribed ACE inhibitors. sions. In many health economies it is the GP and their primary health care team that provide the focus for In a recent study [28] conducted in the Netherlands, this integration, with the other associated health care two procedures for treatment were reviewed by a professionals feeding in their expertise. Good-quality team consisting of a community pharmacist and a computer-based records that can be read by, and re- GP. In one group (the ‘case conference’ group), the ceive input from, all health care professionals provid- pharmacist and GP personally discussed problems as ing support for older people with diabetes are important identified in the pharmacotherapy of the patient, and to facilitate the integration. together drew up a care plan. In the other group (the ‘written feedback’ group), the pharmacist passed the 23.8 Problems of polypharmacy and results of their treatment review back to the GP as written feedback. The effects and cost differences were medication concordance in determined at 6 and 9 months, and the yearly savings older people with diabetes in medicine costs for each year that the medication change persisted were determined. The importance of polypharmacy in older people has In the case conference group, significantly more been recognized in the UK by the Royal College of medication changes were initiated than in the written Physicians (RCP) [27], which confirmed that polyphar- feedback group; this difference was present at 6 months macy affects 16–17% of such patients. Many older after treatment reviews, but was no longer significant people with diabetes will be receiving a number of after 9 months. The authors concluded that perform- different medications for the control of blood glucose, ing treatment reviews with case conferences between blood pressure and lipids. But they may also have other the community pharmacist and GP would lead to a comorbidities such as osteoarthritis, constipation and greater uptake of clinically relevant recommendations respiratory problems, all of which may require drug than would a written feedback, while the extra costs therapy. Hence, the older diabetic patient may well be incurred seemed to be covered by the related savings. receiving six or more medications simultaneously. However, as the effect of the intervention declined over Older patients are more prone to problems related time, the performance of treatment reviews for older to their medicines because of the higher number they people should be integrated into the routine collabo- use, and because of a decline in cognitive and physical ration between GPs and community pharmacists, and functioning [28]. A study of these problems found that should occur at 6-month intervals. two-thirds of all older people had problems using their medicines correctly, and that these problems could lead 23.8.1 Improving medication concordance to a deterioration in a clinical condition for one in four older people [29]. Community pharmacists in a in primary care number of countries are being trained and equipped to The problem of patients not complying with treatment perform treatment reviews on patients taking multiple was first recorded over 2000 years ago, when Hip- therapeutic agents. In some countries, such as the pocrates advised the physician “...to be alert to the UK, community pharmacists may receive a financial faults of the patients which make them lie about their incentive for doing this work [30]. Published studies taking of the medicines prescribed and when things go from the USA [31] and UK [32] have shown that such wrong, refuse to confess that they have not been taking treatment reviews can be helpful. The components of their medicine” [33]. In a retrospective cohort study a treatment review would include: from Scotland [34], a total of 2920 subjects with at least 12 months’ prescriptions for oral hypoglycaemic • A review of the patient’s dug history. agents was identified, and their adherence to treatment • A review of all currently administered medications estimated using data gathered from dispensed prescrip- both prescribed and any bought ‘over-the-counter’. tions. 334 CH 23 PRIMARY AND COMMUNITY CARE OF DIABETES IN OLDER PEOPLE An adequate adherence to treatment – defined adherence perspective, and combination preparations as dispensed doses of at least 90% of doses should also be considered to reduce the tablet load. prescribed – was identified in only 31% of patients • Using treatments that have few, if any, side effects receiving sulphonylurea monotherapy, and 34% with wherever possible. metformin monotherapy. There were significant linear trends of poorer adherence with each increase in daily • Ensuring that treatments are appropriately packaged number of tablets taken, and increase in co-medication and the labelling is clear. Several reasons have been proposed why elderly • Considering the use of a pre-filled tablet dispensing people may not take their tablets as prescribed, includ- system if forgetting whether tablets have been taken, ing: or not, is becoming an issue. • A lack of education or understanding regarding appropriate self-administration and the importance of daily treatment. 23.9 Conclusions • Confusion over which tablets to take, and when; this As the numbers of older people in society with di- may occur especially in older people with developing abetes continues to increase, there will be a shift in memory loss. care from the hospital to the community, and improved • Any changes in the drug or dose regimen. systems of primary care will be required to provide optimal care. Today, there is evidence that multidisci- • Unpleasant side effects. plinary team working, pay-for-performance incentives, • Physical problems opening the packaging, or prob- the individualization of care and dealing with issues lems reading the label. of polypharmacy and medication concordance, may all play their part in helping to deliver optimal to the per- • The demands of a busy lifestyle. son with diabetes in the community. A systematic review of interventions to enhance medication adherence in chronic medical conditions, conducted in 2007 [35], identified 37 eligible trials, in References 12 of which the intervention was informational, in 10 1. Diabetes Atlas (2003), 2nd edition. IDF, Belgium. behavioural, and in 15 combined informational, be- 2. Starfield B 1992 Primary Care. Oxford University havioural and/or social investigations. Press, New York, USA. Twenty of these studies reported a significant im- 3. Forrest CB. Primary care gatekeeping and referrals: provement in at least one adherence measure. Adher- effective filter or failed experiment. Br Med J 2003, ence was increased most consistently with behavioural 326: 692–5. interventions that reduced dosing demands, and those 4. Testing Times: A review of diabetes services in Eng- involving monitoring and feedback. Adherence was land and Wales. The Audit Commission, London, also improved in six multi-sessional information trials 2000. and eight combined interventions. The authors con- 5. Griffin S. Diabetes care in general practice: a cluded that several types of intervention may be ef- meta-analysis of randomised controlled trials. Br Med fective in improving medication adherence in chronic J 1998, 317: 390–6. medical conditions, but few had any significant effect 6. Clinical Guidelines for Type 2 Diabetes: European on the clinical outcome. Diabetes Working Party for Older People 2001-2004 A simple strategy can be developed to try to improve ( www.eugms.org) (last accessed 20 October 2007.) 7. British Medical Association (2003) Investing in Gen- medication concordance in older people with diabetes: eral Practice: The New General Medical Services Con- • Carefully explaining to people what each tablet is tract. British Medical Association, London, UK. Avail- for, when it should be taken, and the importance of able at: www.bma.org. remembering to take each treatment as prescribed. 8. Campbell S, Reeves D, Kontopantelis E et al. Quality of primary care in England with the introduction of pay • Trying to minimize the number of tablets to be for performance. N Engl J Med 2007, 357: 181–90. taken, and the frequency with which they need to be 9. Khunti K, Gadsby R, Millett C et al. (2007) Quality taken. Once-daily tablet treatments are ideal from the of diabetes care in the UK: comparison of published 23.9 CONCLUSIONS 335

quality of care reports with results from the Quality and 22. Sirois C, Moisan J, Poirer P and Gregoire J-P. Sub- Outcomes Framework for Diabetes. Diabetic Medicine optimal use of cardioprotective drugs in newly treated 24, 1436–41. elderly individuals with type 2 diabetes. Diabetes Care 10. United Kingdom Prospective Diabetes Study (UKPDS 2007, 30, 1880–2. 33). (1998) Intensive blood glucose control with 23. Gadsby R. Care of people with diabetes who are house- sulphonylureas or insulin compared with conventional bound or in nursing and residential homes. Diabetes in treatment and risk of complications in patients with General Practice 1994, 4, 30–1. type 2 diabetes. Lancet, 352; 837–53. 24. British Diabetic Association, (1999) Guidelines of 11. Muggeo M, Verlato G, Bonara E, et al. Long-term Practice for Residents with Diabetes in Care Homes. instability of fasting plasma glucose predicts mortality British Diabetic Association, London. in elderly NIDDM patients: the Verona Diabetes Study. 25. Kane RL, Keckhafer G, Flood S et al. The effect of Diabetologia 1995, 38, 672–8. Evercare on hospital use. J Am Geriatr Soc 2003, 51, 12. Kuusisto J, Mykkanen L, Pyorla K and Laakso M. 1427–34. NIDDM and its metabolic control predict coronary 26. Gravelle H, Dusheiko M, Sheaff R et al. Impact of heart disease in elderly subjects. Diabetes 1994, 43, case management (Evercare) on frail elderly patients: 960–7. controlled before and after analysis of quantitative 13. Nathan DM, Singer DE, Godine JE et al. Retinopathy outcome data. Br Med J 2007, 334: 31–4. in older type 2 diabetics: association with poor control. 27. Royal College of Physicians (RCP) (1997) Medications Diabetes 1986, 35, 797–801. for Older people. Royal College of Physicians, Lon- 14. Heart Protection Study Group. (2002) MRC/BHF Heart don. Protection Study of cholesterol lowering with sim- 28. Denneboom W, Dautzenberg MG, Grol R and Smet PA. vastatin in 20,536 high-risk individuals: a randomised Treatment reviews of older people on polypharmacy in placebo-controlled trial. Lancet, 360, 7–22. primary care. Br J Gen Pract, 2007, 57, 723–31. 15. Colhoun HM, Betteridge DJ, Durrington PN et al. 29. Denneboom W, Dautzenberg MG, Grol R and Smet PA. (2004) Prevention of cardiovascular disease with Ator- User-related pharmaceutical care problems and factors vastatin in type 2 diabetes The CARDS multicentre affecting them: the importance of clinical relevance. J randomized, placebo controlled trial. Lancet 2004, 4: Clin Pharm Ther, 2005, 30, 215–23. 685–96. 30. Pharmaceutical Services Committee 2007 (www.psnc. 16. United Kingdom Prospective Diabetes Study (UKPDS org.uk; last accessed 28 November 2007). 38). (1998) Tight blood pressure control and risk of 31. Hanlon IT, Weinberger M, Samsa GP et al.Aran- macrovascular and microvascular complications in type domised controlled trial of a clinical pharmacist inter- 2 diabetes (UKPDS 38). Br Med J, 317; 703–13. vention to improve inappropriate prescribing in elderly 17. Hansson L, Zanchetti A, Carruthers SG et al. (1998) Ef- outpatients with polypharmacy. Am J Med, 1996, 100, fects of intensive blood pressure lowering and low dose 428–37. aspirin therapy in patients with hypertension. Principal 32. Zermansky AG, Petty DR Raynor DK et al.Ran- results of the Hypertension Optimal Treatment (HOT) domised controlled trial of clinical medication review randomised trial. Lancet, 351; 1755–62. by a pharmacist of elderly patients receiving repeat pre- 18. Sinclair AJ, Bayer AJ and the All Wales Research in scriptions in general practice. Br Med J, 2001, 323, Elderly (AWARE) Diabetes Study (1998) UK Govern- 1340–3. ment Report 121/3040, Department of Health, London. 33. Sawyer S. Adherence: Whose Responsibility? http:// 19. Mangin D, Sweeney K and Heath I. Preventive health www.nationalasthma.org.au/html//management/ care in elderly people needs rethinking. Br Med J 2007, adherence/adh009 (accessed 14 December 2008). 335: 285–7. 34. Donnan PT, MacDonald TM, Morris AD for the 20. Bjerre LM and LeLorier J. Do statins cause cancer? A DARTS/MEMO Collaboration. Adherence to oral hy- meta-analysis of large randomised controlled clinical poglycaemic agents prior to insulin therapy in Type 2 trials. Am J Med, 2001, 110: 716–23. diabetes. Diabetic Medicine, 2002, 19, 279–84. 21. Hippisley-Cox J, Pringle M and Carter R. Coronary 35. Kripalani S, Yao X and Haynes B. Interventions to heart disease prevention and age inequalities: the ﬁrst enhance medication adherence in chronic medical con- year of the national service framework for CHD. Br J ditions. Arch Int Med, 2007, 167, 540–50. Group Pract, 2006, 11, 27–31. 24 Diabetes Care in Special Circumstances

I Acute Hospital Admission II Serious Infections Jay Chillala Trafford General Hospital, Manchester, UK III Minor Ethnic Populations Gurch Randhawa University of Bedfordshire, Putteridge Bury, Luton, UK I Acute Hospital Admissions

disorders, depression and osteoporosis, also increase Key messages the likelihood of a hospital admission (see Table 24.1) • Hospital admissions are higher in elderly pa- [2]. Comorbidities increase the length of stay in tients with diabetes. hospital, with the stay doubling if five are present. • Comorbidities increase the length of stay. The presence of nephropathy, coronary/peripheral • Discharge planning can be complex and should artery disease, higher glycated haemoglobin (HbA1c) start early. level, insulin treatment and older age all predict a higher rate of hospitalization [3]. 24.1 Introduction 24.2 Hyperosmolar hyperglycaemic state and diabetic ketoacidosis Patients with diabetes have an increased frequency of hospitalization compared to patients without diabetes. insulin regimes Over 50% of patients with diabetes are admitted at least once for any cause. Compared to patients without The following three conditions are described in greater diabetes, there is an excess risk of 30% [1]. The rates detail elsewhere in this book, and so will be dealt are higher in patients with comorbid conditions such as with only briefly at this point. Treatment consists of congestive cardiac failure, cardiomyopathy, coronary insulin, fluids and potassium. The regimes suggest the atherosclerosis and hypertension (see Table 24.1). use of a syringe pump with 50 units soluble insulin and Non-cardiovascular comorbidities in elderly patients 50 ml saline, starting at 0.1 U/kg/h [4]. A sliding scale with diabetes, such as chronic obstructive pulmonary is subsequently followed, which should be modified disease (COPD), asthma and lower respiratory according to response, and 5% dextrose started when disorders, Alzheimer’s disease/dementia, personality the glucose level is <12 mmol/l. When the patient

Table 24.1 Reasons for acute hospital admissions in patients with diabetes. Acute metabolic complications • Hyperosmolar hyperglycaemic state • Diabetic ketoacidosis • Hypoglycaemia

Poor diabetes control • For monitoring and correction of cause and modiﬁcation of therapy

Infections • Chest, cystitis, pyelonephritis, septicaemia, cellulitis

Cardiac causes • Chronic heart failure, arrhythmias, ischaemic heart disease, hypertension

Ophthalmic complications • Cataract, glaucoma, blindness leading to complications (e.g. falls)

Renal complications • End-stage renal failure

Neurological • Cerebrovascular disease

Tissue diseases • Amputation, foot and lower limb ulcers, pressure sores

is eating and the ketone level is ≤1+ (as per the 24.3 Factors which affect dipstick scale), then subcutaneous insulin treatment can be started. management In diabetic ketoacidosis (DKA), fluid depletion is Older patients may be admitted with conditions that 3–6 litres, and the aim should be replacement over a 24-h period. Care is needed when replacing fluids in occur secondary to their diabetes, such as falls or leg patients with heart failure so as not to fluid overload. ulcers occurring as a consequence of their neuropathy The rapid replacement of fluids may also lead to or retinopathy. Their functional status may then be cerebral oedema; consequently, it is also important to affected by this, leading to higher care requirements monitor levels of consciousness. and possibly a care home on discharge. Elderly diabetic patients in care homes have a higher 24.2.1 Hyperosmolar hyperglycaemic state level of cognitive impairment and an already existing high level of dependency [6]. These factors would (HHS) delay the time for recovery. In HHS, the fluid loss is 8–10 l, and normal saline Other factors affecting management include depres- should be administered until the patient is rehydrated. sion, social problems, poor vision and difficulties with When the blood glucose level is <12 mmol/l, dextrose activities of daily living (see Table 24.2). may be administered instead of saline. If the plasma sodium level is >160 mmol l−1, 0.45% (w/v) saline can be used for the first 3 litres of replacement fluid. 24.4 Discharge planning

Discharge planning should begin at an early stage, 24.2.2 Hypoglycaemia preferably on admission, with information being col- Autonomic symptoms start at glucose levels of lected about care packages and other help received by 3.3–3.6 mmol l−1, and neuroglycopenia at <2.6 mmol/l patients. Both, relatives and patients should be encour- [5]. Autonomic symptoms include sweating, tremor aged to highlight areas where additional input may be and palpitations, with neuroglycopenia presenting as required. confusion and drowsiness leading on to seizures and Multi-disciplinary input on a regular basis is very coma if not treated. Treatment is with 50 ml of 50% important, with involvement of the doctor, physiother- dextrose or intramuscular/intravenous glucagon (1 mg). apist, occupational therapist and social worker. In view 24.4 DISCHARGE PLANNING 339

Table 24.2 Summary of key in-patient criteria for managing older people with diabetes. Management Reason for consideration Assessment of preadmission status Aim to achieve preadmission status at time of discharge Early use of antibiotics Infections can escalate rapidly and be severe. Signs may be subtle Early use of insulin Infections can be severe, blood sugars very variable, good control needed rapidly. Patients may be admitted with HHS. Check renal function regularly Rapid deterioration can occur with some antibiotics. Monitor to check for correction of dehydration. Daily review of fluids and antibiotics Fluid overload/dehydration can occur rapidly. Complications related to antibiotics can occur Review of diabetes control and other Patients may not have attended their annual review modifiable risk factors Review of current medication Check for drug interactions, ability to take medications Early multidisciplinary assessment Important to maintain physical abilities and cognitive function as soon as possible Weekly goal setting Patient condition may improve or decline. Realistic goals need to be set. Early discharge planning Discharge may be complex. Consider timescale and people involved e.g. social services, family Post-discharge review Follow-up of elderly patients with complex problems important. Establish where review will be done and by whom. Possibly outpatients/own home/day hospital. Review can be by doctor/diabetes nurse/community nurse of the comorbidities and general frailty, elderly patients input it can be successful, allowing patients to remain may require a period of rehabilitation, and so the timing at home and reduce the risk of readmission. of discharge becomes very important. The aim should Follow-up and the continuation of care is extremely be to achieve the predischarge status. important, and continues with the GP and consultant, Communication with relatives needs to be effective, with additional input from district nurses and health with discussion about the patient’s needs and the care visitors. Some areas will have a specialist input from packages that social services can provide. Patients may a diabetes nurse specially trained in treatment of the require modifications to their home prior to discharge, elderly, who will monitor blood sugars and visit those or step up to residential/nursing care. These discussions patients in care homes. must be started as early as possible. Discharge planning can be very complicated and time-consuming, but with the appropriate specialist 340 CH 24 DIABETES CARE IN SPECIAL CIRCUMSTANCES II Serious Infections

24.6 How is the treatment of elderly Key messages patients different? • Serious infections in the elderly may present atypically. Clinical signs may be subtle or atypical, with patients • Consequences of infections may be serious due presenting with symptoms such as poor mobility, con- to comorbidities. • fusion or lethargy. This may occur on a background of Care is required for the dosing and the route cognitive impairment, making the diagnosis very chal- used when administering antibiotics. lenging. The usual signs of infection such as fever may not occur, and in fact in severe infections temperatures may drop below baseline. 24.5 Introduction Changes in the immune system make elderly patients more susceptible to bacterial and viral infections. The Serious infections in the elderly frequently present in immune system undergoes changes with age [8]. Due an atypical manner (e.g. falls, incontinence), and with to thymic involution, the T cells that respond to new very few clinical signs. It is therefore crucial that antigens decrease in number (naive T cells), such that investigations are conducted promptly and appropriate most of the T cells are memory T cells that respond treatment instituted at an early stage. to the exposure of previous antigens; however, the The consequences of infections may be serious due response to new antigens is poor. Humoral immunity to the patient’s comorbidities, although the organ- is also mildly affected, but only where the T helper isms may be the same as those encountered in the cells mediate B-cell functions. Immunoglobulin levels younger population. This, combined with immune sys- are constant, although the levels of IgG, IgM and/or tem deﬁcits, may lead to patients being susceptible to IgA may change. life-threatening complications (Table 24.3).

Table 24.3 Common and more serious infections in patients with diabetes [7].

Site of infection Infection Head and neck • Rhinocerebral mucormycosis • ‘Malignant otitis externa’ • Ophthalmic infections

Mouth and oesophagus • Oral candidiasis, • Oesophageal candidiasis

Chest • Community-acquired infections • Pneumococcal and inﬂuenza infections • Tuberculosis, Staphylococcus aureus, Klebsiella pneumoniae and other Gram-negative organisms

Abdomen • Emphysematous cholecystitis, retroperitoneal abscess, hepatic abscess and pelvic abscess

Renal • Bacteruria, cystitis, pyelonephritis, renal/perinephric abscess, emphysematous cystitis/pyelonephritis

Skin, soft tissue and bone • Cellulitis, wound infections, foot ulcers, osteomyelitis, Fournier’s gangrene, necrotising fasciitis/cellulitis 24.8 PRECAUTIONS FOR THE MEDICAL MANAGEMENT OF INFECTIONS IN THE ELDERLY 341

Elderly patients may already have symptoms of fungi can be involved, including Rhizopus, Absidia illnesses, which makes the diagnosis very difficult. and Mucor) into the paranasal sinuses. From there it Many patients have incontinence or an indwelling may spread to the palate, sphenoid sinus, cavernous catheter that may mask the signs of infection. Although sinus, orbits and the brain. Presentation can be by asymptomatic bacteriuria is very common, it should fever, confusion, a visible black eschar or cranial not be treated unless the symptoms of a urinary tract nerve palsies. Treatment is with amphotericin B and infection develop. debridement. Elderly patients with diabetes are more prone to • certain infections, especially skin infections. They are Malignant otitis externa: This is an invasive infec- also likely to have peripheral vascular disease and tion of the external auditory canal and skull base. neuropathies, which makes ulcers and skin lesions Treatment is with a quinolone. more likely. • Emphysematous cystitis/pyelonephritis: Patients Patients may already be taking several medications, present with symptoms of a urinary tract infection and care is needed to ensure that any antibiotic treat- with X-rays showing the presence of air in the ment does not interact with existing tablet regimes. It bladder lumen, bladder wall, the renal parenchyma is also important to ensure that drug levels are kept or perinephric space. Treatment for the former is within limits, especially when renal impairment exists. with antibiotics, and the latter with antibiotics and relief of the obstruction. In some cases nephrectomy 24.7 Management is necessary. The antibiotics of choice would be cephalosporin or quinolone. Treatment is along standard lines, although classical signs are often not seen and antibiotics are used at • Fournier gangrene: the male genitalia are affected an early stage with therapy often given empirically. by necrotizing fasciitis. Treatment is debridement, Details of the suggested management are as follows: combined with antibiotics active against Streptococ- cus sp., Clostridium sp. and Gram-negative organ- • Oropharyngeal infections isms. ◦ Oral candidiasis: nystatin or oral fluconazole ◦ Oesophageal candidiasis: oral fluconazole • Necrotizing cellulitis/fasciitis: treat with debridement and antibiotics. • Chest infections • Streptococcus pneumoniae: penicillin or a macrolide if penicillin-allergic • Gram-negative bacilli: Klebsiella/Pseudomonas 24.8 Precautions for the medical aeruginosa: third-generation cephalosporin plus management of infections in macrolide the elderly • Legionella pneumophilia: macrolide ± rifampicin • Chlamydia and Mycoplasma: macrolide Elderly patients with an infection can deteriorate • Influenza/parainfluenza/RSV: oseltamivir and rapidly, and attempts should be made to identify the zanamivir can be given within 48 h of onset of source of sepsis with blood, urine cultures or wound symptoms swabs. Broad-spectrum antibiotics should be started at The following infections are more common in pa- an early stage, with the appropriate changes when or- tients with diabetes: ganism sensitivities are available. The renal function and glomerular filtration rate • Tuberculosis decreases with age; consequently, the dose levels of • Klebsiella drugs that are normally excreted via the renal route will need to be adjusted. • Staphylococcus aureus For example, the aminoglycosides are adjusted ac- 24.7.1 Rare infections (see Table 24.4) cording to the patient’s renal function and their body weight. In patients with a normal renal function, • Rhinocerebral mucormycosis [10]: This infection oc- the dose levels should be monitored after three to curs due to inhalation of the fungus (several different four doses and adjustments made accordingly to the 342 CH 24 DIABETES CARE IN SPECIAL CIRCUMSTANCES

Table 24.4 Summary of treatment of infections in elderly diabetes patients. (Adapted in part from Ref. [9]). Head and neck • Rhinocerebral mucormycosis • Surgical debridement and amphotericin B • Malignant otitis externa • Quinolone

Mouth and oesophagus • Oral candidiasis • Nystatin or ﬂuconazole • Oesophageal candidiasis • Fluconazole

Chest • Community-acquired infections • Amoxicillin or a macrolide if penicillin allergic • Atypical pathogen • Erythromycin or clarithromycin • Severe Legionella • Add rifampicin • Staphylococcal e.g. in inﬂuenza • Add ﬂucloxacillin • Staphylococcal, if MRSA suspected • Add vancomycin • Gram-negative bacilli • Cefuroxime (or cefotaxime) plus macrolide

Abdomen • Emphysematous cholecystitis, • Tazocin, or aminoglycoside/quinolone with clindamycin or metronidazole • Hepatic abscess and pelvic abscess • Tazocin or third-generation cephalosporin, or ﬂuro- quinolone plus metronidazole. Antibiotics given whilst awaiting results of Gram stain and culture Renal • Bacteruria, cystitis, pyelonephritis, renal/perinephric • Cephalosporin or quinolone abscess, emphysematous cystitis/pyelonephritis

Skin, soft tissue, bone • Necrotizing cellulitis/fasciitis • Surgical debridement and broad spectrum antibiotics (e.g. penicillin, clindamycin) • Fournier’s gangrene • Treatment similar to above, addition of quinolone or aminoglycoside plus surgical debridement dosage. In those patients with a once-daily regime, lo- If tablets are large and difﬁcult to swallow, liquid cal guidelines on monitoring drug serum levels should preparations may provide an alternative, although the be followed. Aminoglycosides should also (preferably) doses may be less accurate. All tablet bottles must not be co-administered with ototoxic diuretics (e.g. be accessible and easily opened; if necessary dosette furosemide), although if this is unavoidable then the boxes may need to be issued. dose interval been the two drugs must be as long as possible. 24.9 When should the The route of drug administration must also be con- microbiologist be called? sidered, with the oral route being ideal (if possible). If a patient is confused, or compliance at home is likely Microbiologists should be contacted at an early stage to be poor, they may require to be admitted to hospital when diabetic patients have complicated or rare infec- for intravenous treatment. tions. If a patient’s condition is deteriorating despite 24.9 WHEN SHOULD THE MICROBIOLOGIST BE CALLED? 343 appropriate antibiotics, it is useful to discuss changes microbiologist may well wish to be involved closely in antibiotics. There may also be uncertainty regard- in the dosing and monitoring of these patients. ing the source of infection, and the microbiologist can Serious infections represent a major challenge, and suggest the best broad-spectrum regime. especially so in the elderly diabetic patient. Hence, it The serum levels of antibiotics such as the amino- is important to be aware of their atypical presentations, glycosides have a narrow therapeutic margin, and the and take rapid actions. 344 CH 24 DIABETES CARE IN SPECIAL CIRCUMSTANCES III Minority Ethnic Populations

and African-Caribbean populations in the UK, and Key messages the disproportionately higher numbers of South Asians • The increased rate of type 2 diabetes among and African-Caribbeans progressing towards diabetic minority ethnic groups compared to Caucasians nephropathy. in the UK is well documented. According to the Census, 4.6 million people are • Diabetes complications, such as end-stage re- from minority ethnic groups. This represents a total nal failure, are much more prevalent among of 7.9% of the total population of the UK. Some 75% the UK’s South Asian and African-Caribbean of the minority ethnic populace are classified as either population. Black/Black British (24.8%) or Asian or Asian British • Inequalities currently exist in diabetes and renal (50.2%) [12]. services; the solutions to rectifying this situation Diabetes is becoming one of the greatest health are complex, and involve focusing both upon problems facing the UK today. The recent ‘All Par- disease prevention and disease management. liamentary Group’ for Diabetes and Diabetes UK re- • The financial and human burden of not address- ported that over three million people are expected to ing these inequalities should encourage some be diagnosed with diabetes by the year 2010, and that immediate action. half of these cases will be people from disadvantaged communities. According to the Diabetes National Ser- vice Framework, people of South Asian, African and 24.10 Introduction African-Caribbean descent have a higher than average risk of developing type 2 diabetes, as compared with The UK Government’s Cross Cutting Review of Health the white population [13]. Type 2 diabetes is up to Inequalities published earlier this decade reminded us sixfold more common in people of South Asian de- not only that health gaps still exist in the UK but also, scent, and up to threefold more common in those of in some cases, they are growing ever wider: African-Caribbean descent. It has been estimated [14] “There are wide geographical variations in health sta- that 15.2% of the South Asian population had diabetes, tus, reflecting the multiple problems of material disad- as compared to 3.8% of the white population. The risk vantage facing some communities. These differences of death resulting from the complications of diabetes begin at conception and continue throughout life. Ba- is between three- and sixfold higher within minority bies born to poorer families are more likely to be born ethnic groups [14]. prematurely, are at greater risk of infant mortality and have a greater likelihood of poverty, impaired development and chronic disease in later life. This sets up an 24.11 Epidemiology of diabetes inter-generational cycle of health inequalities.” [11]. among minority ethnic This statement reflects the shift in the focus of policy groups in the UK during the past 20 years, in which there has been a growing interest in the health of minority ethnic popu- Both, South Asian and African-Caribbean communities lations in the UK. (For the purposes of this chapter and have a high prevalence of type 2 diabetes, with recent availability of data, the term ‘minority ethnic groups’ studies having indicated a prevalence rate fourfold is used to refer to South Asian and African-Caribbean greater than that in Whites (Table 24.5). It has been populations in the UK.) reported that 20% of South Asians aged 40–49 years Throughout this period, the provision of diabetes have type 2 diabetes, and that by the age of 65 services for minority ethnic groups has become a par- years that proportion will have risen to one-third [15]. ticularly important area of debate. This is in part due to Clearly, this has significant consequences for planning the observation of high rates of type 2 diabetes among diabetes services for the elderly in areas of the UK South Asian (those originating from the Indian sub- where there are higher proportions of minority ethnic continent, India, Sri Lanka, Pakistan and Bangladesh) groups. 24.13 LOOKING TO THE FUTURE 345

Table 24.5 Relative risk of dia- and its complications is poor among South Asians betes and diabetic nephropathy among and African-Caribbeans [20, 21]. Preliminary evidence South Asians and African-Caribbeans also suggests that the quality of health care for South in the UK population. Asians and African-Caribbeans is inadequate and the Condition Relative risk compliance poor [15, 21]. There is also a low-uptake of hospital-based diabetes services, with growing ev- Diabetes >4 idence that South Asians are subsequently referred Diabetic nephropathy >6 later for renal care, and are more likely to be lost to follow-up [22]. Late referral may reduce opportunities to implement measures to slow the progression of renal A further complication is that diabetic nephropathy failure, or to prepare adequately for RRT, thus adding is the major cause of end-stage renal failure (ESRF) to both morbidity and mortality. in South Asian and African-Caribbean patients receiv- The World Health Organization (WHO) study group ing renal replacement therapy (RRT), either by dialysis on diabetes has noted that resources should be directed or transplantation. Nationally, this higher relative risk, towards improving the quality of preventive care in when corrected for age and gender, has been calculated primary care settings and to public health interventions in England as 4.2 for the South Asian community and for controlling diabetes. Education, early diagnosis 3.7 for those with an African-Caribbean background and effective management of diabetes is important [16]. Data acquired from Leicester show that South for safeguarding the health of susceptible populations Asians with diabetes have a 13-fold higher risk of de- and for long-term savings for the NHS [15]. Most veloping ESRF than do ‘White’ Caucasians [17]. Thus, encouragingly, recent studies from the United States not only are South Asians and African-Caribbeans and Finland have demonstrated that modest lifestyle more prone to diabetes than Whites, they are more changes can reduce the risk, by more than 58%, of likely to develop ESRF as a consequence. developing overt type 2 diabetes in susceptible groups Importantly, the South Asian and African-Caribbean [23, 24]. Furthermore, various interventions, such as populations in the UK are relatively young compared tight blood pressure control, and the effective use to the White population. Since the prevalence of ESRF of angiotensin-converting enzyme (ACE) inhibitors or increases with age, this has major implications for angiotensin receptor (ATR) blockers and tight blood the future need for RRT, and highlights the urgent sugar control can significantly delay the progression need for preventive measures [18]. The incidence of of diabetic nephropathy [25–30]. ESRF has significant consequences for both local and national NHS resources. The National Renal Review has estimated an increase over the next decade of 80% 24.13 Looking to the future in the 20 000 or so patients receiving RRT, and a doubling of the current cost – to about £600 million It is clear that minority ethnic groups are disproportion- per year – of providing renal services [15]. ately affected by diabetes and consequent renal health Consequently, there is an urgent need to invest in problems, both in terms of access to appropriate ser- renal services as well as diabetes services, given the vices and the higher prevalence of diabetes and renal greater propensity of diabetes complications among complications. diabetics from minority ethnic groups. A major undertaking for researchers and clinicians in the UK will be to explore access to and the progression through the diabetes and ‘renal disease compli- 24.12 Improving access to services cating diabetes’ care pathways, and to identify health beliefs and experiences associated with diabetes and The Diabetes National Service Framework highlights diabetic renal complications among African-Caribbean the importance of access to services, in particular to and South Asian groups. A systematic exploration of meet the needs of minority ethnic groups [13]. The these would provide a valuable resource for health pro- Renal Services NSF also focuses on ‘renal disease fessionals working with these groups, and allow for the complicating diabetes’, and emphasizes inequalities development of a culturally competent diabetic and re- experienced by minority ethnic groups [19]. How- nal service, which is sensitive to the needs of minority ever, there is evidence that knowledge of diabetes ethnic groups [31]. 346 CH 24 DIABETES CARE IN SPECIAL CIRCUMSTANCES Specifically, these gaps are: incidence of diabetes among minority ethnic groups via preventive strategies. • The identification of cultural beliefs and practices There has been substantial recognition of the need to relevant to diabetes and diabetic renal disease improve diabetes care among minority ethnic groups, self-management, including attitudes to medication as evidenced by a number of new initiatives launched and attendance to GPs, diabetic services and by the NHS National Diabetes Support Team and nephrology services for routine monitoring. Diabetes UK. Many of these initiatives are recognized • The examination of referral patterns to to be at the forefront worldwide in the development hospital-based diabetic services, and subsequent of culturally competent diabetes education materials. attendance. However, the success of these initiatives has been limited by the lack of a focused strategy that brings • The exploration of referral patterns to nephrology together the various strands of a multi-faceted problem services. that would lead to a coherent implementation plan. It • The exploration of the relevance of current renal is hoped that this chapter will contribute to beginning and shaping this process, not only in the UK but also complications education programmes for minority for many other countries who have a multi-ethnic and ethnic groups [32]. multi-faith society. Kidney Research UK have recently launched On a final note, it is worth remembering that the the ABLE – “A better life through education and relatively young minority ethnic population ages, the empowerment” – campaign, which aims to redress burden of diabetes and its complications will become some of the above issues by education and raising unsustainable in many parts of the UK, given the personal awareness of kidney health issues among disproportionate incidence of diabetes among these minority ethnic groups. Professor Randhawa is leading communities. It is also worth noting that detailed a national pilot study (with colleagues from Imperial debates concerning diabetes and ethnicity are relatively College and the University of Leicester) to explore the new, and are limited by the quality of data available above issues, and the study is to due to be completed not only in the UK but also worldwide. In future, in 2009. it is imperative that data are collected on a wide Concomitantly, there needs to be a reduction in the range of variables including age, ethnicity, social class, incidence of diabetes among the African-Caribbean gender and religion. The potential interaction of these and South Asian population in order to alleviate the hu- variables will be an important area of research in man and economic costs of this condition. This process future to identify more effective methods in targeting can only begin if the public are in an informed position resources to those populations most in need [33]. to consider and debates the issues of disease preven- It is only when these issues are addressed adequately tion, healthy living and self-management. Central to that a diabetes service might emerge that can truly meet attaining this goal are increased levels of health educa- the needs of a multi-ethnic and multi-faith population tion and awareness of the specific problems within the within the UK. African-Caribbean and South Asian population. This is a difficult challenge as many of these communities live within the most deprived (and difficult-to-reach) References communities in the UK [12]. 1. Bo S, Ciccone G, Grassi G, Gancia R and Rosato R. Patients with type 2 diabetes had higher rates of 24.14 Conclusions hospitalization than the general population. J Clin Epi- demiology 2004, 57 (11), 1196–201. 2. Aro S, Kangas T, Reunanen A, Salinto M and Kovisto Inequalities do currently exist in diabetes services in V. Hospital use among diabetic patients and the general the UK, and the solutions to rectifying this situation are population. Diabetes Care 1994, 17 (11), 1320–9. complex. They also require an holistic approach that 3. Rosenthal MJ, Fajardo M, Gilmore S, Morley JE and considers both the short-term requirement to manage Nailboff BD. Hospitalization and mortality of diabetes diabetes and its complications among minority ethnic in older adults. A 3 year prospective study. Diabetes groups, and the longer-term focus to decrease the Care 1998, 21 (2), 231–5. 24.14 CONCLUSIONS 347

4. Lebovitz HE. Diabetic ketoacidosis. Lancet 1995, 345, 22. Jeffrey RF, Woodrow G, Mahler J, Johnson R and 767–72. Newstead CG. (2002) Indo-Asian experience of renal 5. Kearney T and Dang C. Diabetic and endocrine emer- transplantation in Yorkshire: results of a 10year survey. gencies. Postgrad Med J 2007, 83, 79–86. Transplantation; 73: 1652–7. 6. Sinclair AJ, Allard I and Bayer A. Observations of 23. Diabetes Prevention Program Research Group (DP- diabetes care in long-term institutional settings with PRG). (2002) Reduction in the incidence of Type 2 measures of cognitive function and dependency. Dia- diabetes with lifestyle intervention or metformin. N betes Care 1997, 20 (5), 778–84. Engl J Med, 346: 393–403. 7. Calvet CM and Yoshikawa TT. Infections in diabetics. 24. Tuomilehto J, Lindstrom¨ J, Eriksson JG, Valle TT, Infect Dis Clin North Am 2001; 15: 407–21. Ham¨ al¨ ainen¨ H, Ilanne-Parikka P, Keinanen-¨ Kiukaan- 8. Weksler ME. Senescence of the immune system. Med niemi S, Laakso M, Louheranta A, Rastas M, Salminen Clin North Am 1983; 67: 263–72. V and Uusitupa M; Finnish Diabetes Prevention Study 9. Rajagopalan S. Serious infections in elderly patients Group. (2001) Prevention of type 2 diabetes mellitus with diabetes mellitus. Clinical Infectious Diseases by changes in lifestyle among subjects with impaired 2005; 40: 990–6. glucose tolerance. N Engl J Med, 344: 1343–50. 10. Tierney MR and Baker AS. Infections of the head and 25. UK Prospective Diabetes Study (UKPDS) Group. neck in diabetes mellitus. Infect Dis Clin North Am (1998) Intensive blood-glucose control with sulpho- 1995; 9: 195–216. nylureas or insulin compared with conventional 11. Department of Health (2002) Tackling health inequali- treatment and risk of complications in patients with ties: Cross Cutting Review. Department of Health, Lon- type 2 diabetes (UKPDS 33). Lancet, 352: 837–53. don. 26. Feest, T.G., Dunn, E.J. and Burton, C.J. (1999) Can 12. Randhawa G (2008) Organ donation and intensive treatment alter the progress of established transplantation – The realities for minority eth- diabetic nephropathy to end-stage renal failure? QJ nic groups in the UK. In: W. Weimar, M.A. Bos, Med, 92: 275–82. J.J. van Busschbach (eds). Organ Transplantation: 27. Brenner BM, Cooper ME, de Zeeuw D, Keane WF, Ethical, Legal and Psychosocial Aspects. Towards Mitch WE, Parving HH, Remuzzi G, Snapinn SM, a Common European Policy. Pabst Publishers, Zhang Z and Shahinfar S; RENAAL Study Investi- Lengerich, Germany. gators. (2001) Effects of losartan on renal and cardio- 13. Department of Health. (2002) National Service Frame- vascular outcomes in patients with type 2 diabetes and work for Diabetes: Standards. Department of Health, nephropathy. N Engl J Med, 345: 861–9. London. 28. Cinotti, G.A. and Zucchelli, P.C. (2001) Effect of 14. Burden, A.C. (2001) Diabetes in Indo-Asian people. lisinopril on the progression of renal insufﬁciency in The Practitioner: 245 (1622): 445–51. mild proteinuric non-diabetic nephropathies. Nephrol 15. Raleigh VS. (1997) Diabetes and hypertension in Dial Transplant, 16: 961–6. Britain’s ethnic minorities: implications for the future 29. Lewis EJ, Hunsicker LG, Clarke WR, Berl T, Pohl of renal services. Br Med J; 314: 209–12. MA, Lewis JB, Ritz E, Atkins RC, Rohde R and Raz 16. Roderick PJ, Raleigh VS, Hallam L and Mallick NP. I; Collaborative Study Group (2001) Renoprotective (1996) The need and demand for renal replacement effect of the angiotensin-receptor antagonist irbesartan therapy amongst ethnic minorities in England. JEpi- in patients with nephropathy due to type 2 diabetes. N demiol Community Health; 50: 334–9. Engl J Med, 345: 851–60. 17. Burden AC, McNally PG, Feehally J and Walls J. 30. Lightstone, L. (2001) Preventing kidney disease: The (1992) Increased incidence of end-stage renal failure ethnic challenge. National Kidney Research Fund, Pe- secondary to diabetes mellitus in Asian ethnic groups terborough, UK. in the United Kingdom. Diabetic Medicine; 9: 641–5. 31. Randhawa, G. (2000) Increasing the donor supply from 18. Randhawa G. (1998) The impending kidney transplant the UK’s Asian Population: the need for further re- crisis for the Asian population in the UK. Public search. Transplantation Proceedings, 32: 1561–62. Health, 112: 265–8. 32. Randhawa, G. (2003) Developing culturally competent 19. Department of Health. (2004) National Service Frame- renal services in the United Kingdom: Tackling in- work for Renal Services. Department of Health, Lon- equalities in health. Transplantation Proceedings, 35: don. 21–3. 20. Nazroo, J.Y. (1997) The Health of Britain’s Ethnic 33. Randhawa, G. (2007) Tackling health inequalities for Minorities. Policy Studies Institute, London. minority ethnic groups: Challenges and Opportunities. 21. Johnson M., Owen D. and Blackburn, C. (2000) Better Health Brieﬁng Paper 6. Race Equality Founda- Black and minority ethnic groups in England: The tion, London. second health and lifestyles survey. Health Education Authority, London. 25 Managing Surgery in the Elderly Diabetic Patient

Geoffrey Gill and Susan Benbow Department of Diabetes and Endocrinology, Aintree University Hospitals, Liverpool, UK

practice is also changing in many countries, with an Key messages increasing number of day-case procedures and shorter postoperative hospital stays. Diabetes management in • Diabetes per se is in general not a barrier to the elderly is also changing with the increasing use of surgery in the elderly person. insulin, and sometimes with more complicated multi- • A thorough preoperative assessment is an es- ple injection regimens and even occasionally the use sential prerequisite in the management of older of insulin pumps [1]. people about to undergo surgery. Although carefully planned and executed surgery • Excessively high blood glucose levels should be is highly successful in the elderly, such patients with avoided during surgery; hypoglycaemia must be avoided at ALL costs. diabetes may tolerate metabolic and infective compli- • Locally agreed protocols for the treatment of cations less well than younger subjects. Diabetes per diabetes during surgery should be available in se should never be a reason to decide not to operate every hospital. on an elderly patient, but it is a reason for careful planning and management – whether preoperatively, perioperatively or postoperatively. In this chapter the 25.1 Introduction potential problems, the basis of current management systems, and the practical methods of treatment will During their lifetime, most patients with diabetes will be examined. require some form of surgery, and the likelihood increases as age advances. Nowadays, a considerable amount of major surgery is undertaken in the elderly 25.2 Metabolic and other problems (e.g. coronary artery bypass grafts, peripheral vascular induced by surgery and aneurysm surgery, removal of malignancies), of whom more are proportionately likely to have diabetes Anxiety, anaesthetic drugs and possibly also the under- than at the earlier stages of their lives. Even during lying condition requiring surgery, may all contribute the past few years in England, there has been a 16% to metabolic destabilization in the diabetic surgical pa- increase in coronary artery bypass grafts and a similar tient. The most important factors, however, are starva- increase in hip replacements in the elderly. Surgical tion and the pathophysiological metabolic and humoral

DIABETIC PATIENT

SURGERY

Pre-existing Interrupted carbohydrate insulin Trauma deficiency intake

Tissue damage, Afferent nerve stimuli Insulin deficiency + insulin antagonism Hypersecretion of: • ACTH and cortisol • Catecholamines • Growth hormone • Glucagon CATABOLIC DRIVE

Gluconeogenesis Hyperglycaemia Glycogenolysis Ketotic tendency Lipolysis

Figure 25.1 Diagram outlining the hormonal and metabolic effects of surgery in the diabetic patient. response to trauma. All but the most minor of opera- afferent nerves from the injured tissue; cortisol, for tions involve some interruption of normal food intake, example, is secreted secondarily to the release of and this may not infrequently last for several days. adrenocorticoptrophic hormone (ACTH) [5]. The re- This poses obvious practical difficulties for diabetic sult of these changes is a massive catabolic drive patients whose tablets or insulin injections must be ac- (see Figure 25.1), with increased gluconeogenesis and companied by the intake of food. Of a more sinister glycogenolysis leading to glucose release into the cir- note, however, is that starvation leads to catabolism culation. Lipolysis and protein breakdown also occur, and, in the presence of insulin deficiency – that is, the although in the non-diabetic even small amounts of diabetic state – ketosis becomes likely and eventually insulin (‘basal’) secretion are sufficient to contain dan- inevitable [2, 3]. gerous hyperglycaemia and lipolysis. This of course Such problems are greatly enhanced by the well- is not true in the insulin-deficient or diabetic state. known humoral and metabolic changes associated with The danger of this metabolic scenario to the diabetic trauma, which also greatly enhance catabolism. Sur- depends on its degree (as mentioned above, this is gical trauma disturbs the usual fine balance between roughly proportional to the severity of the trauma), and anabolism (which effectively is controlled only by to the level of insulin reserves available. insulin) and catabolism (which is driven by a variety of hormones, notably cortisol, catecholamines, growth hormone and glucagon). These latter hor- 25.3 Implications for management mones are often collectively known as the ‘stress’ of surgery in diabetic patients or ‘counter-regulatory’ hormones, and they are hyper- secreted in traumatic states. Cortisol and adrenaline These basic principles can be translated logically into levels in particular rise promptly (within minutes or principles of management for diabetic patients under- hours after the initiation of trauma), and often mas- going surgery [6]. The major requirement is to ensure sively (to some extent in proportion to the degree of adequate insulinization, with the important variables trauma). In addition, insulin secretion is relatively re- being the degree of surgical trauma and the individual duced, and a state of insulin resistance ensues [2–4]. level of endogenous insulin reserves. In practical terms, Many of these changes are neurally mediated via patients can be grouped into those receiving insulin 25.4 POTENTIAL RISKS OF SURGERY IN DIABETIC PATIENTS 351

Diabetic patient requiring surgery

Patient on insulin treatment Patient on diet and/oral agents

Assume little or no insulin Assume patient haslimited reserves insulin reserves

Insulin/glucose infusion Observation only for most needed for all degrees of surgery – if major, use surgery insulin/glucose

Figure 25.2 Schematic flow chart demonstrating the principles of managing diabetes during surgery. treatment, and those treated with dietary control and/or recent study, Hjortrup et al. [8] used modern treatment oral hypoglycaemic agents (OHAs). The insulin-treated methods but failed to show any difference in mortality group may or may not be truly insulin-dependent (type between diabetic and non-diabetic subjects (2.2% ver- 1 diabetes), but even if they are insulin-treated type 2 sus 2.7%). Some specific surgical procedures – notably patients they can be assumed to have little or no in- vascular – may have an increased risk in diabetic pa- sulin reserves, and deemed to require such treatment. tients. Thus, aortic and lower-limb revascularization In fact, these patients require continuous exogenous procedures carry an increased mortality risk in diabetic insulin treatment for all types of surgery. compared to non-diabetic patients [9]. However, this Those not receiving insulin must have at least lim- was clearly a selected diabetic group with established ited insulin reserves, and for minor to moderate degrees advanced large-vessel disease, and such an outcome of surgical trauma can usually be monitored simply by difference may therefore not be surprising. Neverthe- close observation. Major surgery, however, will require less, such results are of relevance to surgery in the continuous insulin (as for the first group above). These elderly, as in this study the mean age of the diabetic principles are summarized in Figure 25.2. group was 68 years. More recently, no increased cardiac morbidity or overall mortality was reported in 25.4 Potential risks of surgery in a similarly aged group of patients with diabetes un- diabetic patients dergoing major vascular surgery when compared to a non-diabetic group [10]. Surprisingly few studies have been conducted to in- On the subject of morbidity, there is little conclusive vestigate postoperative mortality and morbidity, com- evidence available that diabetes per se causes any in- paring diabetic with non-diabetic subjects. Diabetes creased risk. Diabetic patients with pre-existing cardiac has certainly been considered a major risk factor for or renal problems may have increased morbidity, but surgery during past decades. An American study con- not if relatively uncomplicated and properly managed ducted in 1963 reported a 5% postoperative mortality [11, 12]. The results of a recent study showed a worse in a large (n = 487) group of surgical diabetic patients, outcome for patients with diabetes following hip frac- the major causes of death being ketoacidosis, infection ture [13], though this was not due to the diabetes itself and myocardial infarction [7]. It is likely, however, that but rather to a reduced functional status prior to the the methods of management were highly suboptimal fracture, such as stroke. This is an important point, as compared to modern management principles. In a more there appears to be an increased risk of hip fracture in 352 CH 25 MANAGING SURGERY IN THE ELDERLY DIABETIC PATIENT the older person with diabetes [14]. Neither did the risk balance, remembering that they may well present dif- of postoperative infection appear definitely increased, ferently from a younger patient; for example, in the in contrast to normally accepted clinical dogma [8, 12]. elderly patient constipation or even a simple urinary In the Danish study conducted by Hjortrup et al.[8], tract infection may cause confusion. the wound infection rate was identical among both di- The potential iatrogenic complications of diabetes abetic (13/224; 5.8%) and non-diabetic (12/224; 5.4%) management during surgery will be discussed at a later patients. A recent study of diabetic patients undergo- point. Elderly diabetic patients tolerate hypoglycaemia ing coronary artery surgery showed an increased risk poorly [16], and are less efficient at maintaining water of post-surgical infections, but that was accounted for homeostasis than their younger counterparts [17]; con- by excessive postoperative hyperglycaemia [15]. sequently the risk of fluid and electrolyte imbalance is Overall, a critical assessment of the available liter- increased during the postoperative period. Renal im- ature does not support a generally increased risk for pairment can be precipitated or exacerbated by these diabetic patients undergoing surgery, in terms of both changes, while hydration may already be compromised mortality and postoperative complications. by drugs, vomiting, preparation for the operation (e.g. bowel clearance), preoperative starvation, or simply by an inability to obtain or reach fluids. 25.5 Special problems in the elderly

Increased age is known to increase postoperative mor- 25.6 New diabetes therapies and bidity, and possibility also mortality, in general. Whilst surgical management this includes diabetics, again there is no convincing evidence that the effect is significantly greater among The past 5–10 years has witnessed a rapid and ex- such patients. In general, however, diabetic surgical tensive increase in therapies for both type 1 and type patients are frequently older and ‘sicker’ than their 2 diabetes. Some discussion of these treatments are non-diabetic counterparts [12] (typically amputees, necessary at this point, as they may lead to concerns coronary bypass surgery, etc.), but when these fac- regarding both preoperative and perioperative manage- tors are taken into account any increased morbidity ment of the diabetic patient. amongst diabetics becomes insignificant, or at least much less significant. 25.6.1 Insulins When preparing the elderly for surgery, any preoperative assessment should be particularly thorough Analogue insulins are non-human synthetic insulins because of comorbidity and polypharmacy. The pa- which have more favourable absorption characteris- tient may not be able to provide an accurate history tics than the traditional insulins. They are currently in because of memory problems or communication dif- widespread use, and include short-acting (e.g. Lispro, ficulties. Likewise, ischaemic heart disease may be Aspart), long-acting (Glargine and Detemir) and pre- underestimated as the patient may not provide a typ- mixed formulations (e.g. Humalog 25, Novomix 30). ical history of chest pain on exertion if their level of Those patients receiving the short-acting or premixed exercise is limited by another pathology such as os- analogues pose no problem for surgical management, teoarthritis. Pressure area management is important in as the short-acting component does not have a greatly the elderly throughout the period of immobility, and altered time course of action. However, the long-acting particularly so if the individual has diabetes, where analogues are often administered during in the evening, peripheral vascular disease and peripheral neuropathy and their theoretical 24-h action could mean that, when both increase the risks. The nutritional status can al- given the night before surgery their absorption may ready be compromised in the elderly hospitalized pa- continue into the next day, and potentially affect the patient, and further exacerbated by surgery. Prophylaxis tient’s perioperative insulin requirements. One option against venous thromboembolic disease must also be would be to move the Glargine or Detemir injection to considered. An assessment of pain during the postop- earlier in the day prior to surgery, or alternatively sim- erative period can again be difficult in the elderly, due ply to be aware of the problem and (depending on the to communication problems or secondary to demen- dose level) possibly reduce the operative insulin deliv- tia. Postoperative confusion thus requires a thorough ery. In fact, a prolonged absorption of these insulins assessment of the whole patient, their drugs and fluid is not always the case and there is evidence that, for 25.7 THE PRACTICAL MANAGEMENT OF SURGERY IN DIABETIC PATIENTS 353 both Glargine [18] and Detemir [19], twice-daily ad- modern management there should nowadays be little or ministration provides smoother absorption profiles than no excess mortality or postsurgical infection risk. Ide- once-daily. So, the next day ‘hangover’ effect may be ally, the period of hospitalization should not be unduly an exaggerated problem. prolonged, although admission a day or two earlier than Although inhaled insulin was introduced in 2006, usual is often required (see later). Postoperative ke- its use has been limited and the manufacturers recently toacidosis should no longer occur, but hypoglycaemia withdrew this formulation for commercial reasons [20]. is always a risk with intravenous insulin delivery. The Other companies, however, are planning its reintroduc- avoidance of hypoglycaemia is very important – the tion, and the preparation may in time return to use. surgical patient may be unable to perceive or report Inhaled insulin has absorption characteristics similar hypoglycaemia, and low blood glucose levels may to those of subcutaneous short-acting insulin, and does therefore be allowed to become profound and seri- not therefore have implications for surgical manage- ous before detection and treatment. Elderly diabetic ment. patients in general may present atypically with hypoglycaemia and, for instance, appear to be confused. 25.6.2 Type 2 diabetes preparations Additionally, there is no evidence that over-zealous at- The glitazones – rosiglitazone and pioglitazone – are tempts at achieving normoglycaemia are of benefit in insulin receptor sensitizers were introduced during the the surgical situation. Indeed, it was found, paradoxi- recent past and are now widely used, most commonly cally [8], that patients with particularly ‘good’ control in combination with either sulphonylureas or met- appeared to be at greater risk of postoperative com- formin, but increasingly as ‘triple therapy’ with both plications, although this has not been confirmed by drugs [21]. For patients established on these drugs, others [15]. there are no implications for surgery other than those Over recent years, there has been a concerted move which apply to other oral agents. That is, unless glu- to tightly control hyperglycaemia in critically ill dia- cose/insulin infusions are required the drugs should be betic and non-diabetic surgical patients [24]. However, stopped on the day of surgery and restarted when food this is not without some controversy [25, 26], and has intake is resumed. been associated with troublesome hypoglycaemia. Two novel agents for type 2 diabetes treatment Glycaemic aims during surgery should, therefore, be have been recently introduced. Exanetide is an incretin to avoid hypoglycaemia at all costs, but in addition to mimetic, or a GLP-1 (glucagon-like peptide 1) ag- not allow excessive hyperglycaemia, nor to risk ke- onist. It is administered twice-daily, subcutaneously, toacidosis. In numerical terms, plasma glucose levels −1 and lowers the blood glucose level without causing in the region of 6.0–12.0 mmol l would be a reason- −1 weight gain. The mode of action of exanetide is to able compromise target, although below 10 mmol l slow gastric emptying, enhance glucose-dependent in- is preferable. sulin secretion, and suppress glucagon secretion [22]. Sitagliptin inhibits DDP-4, the enzyme involved in 25.7.2 Preoperative assessment the metabolism of incretin hormones such as GLP-1. The preoperative assessment of elderly diabetic pa- Consequently, sitagliptin has similar effects to exane- tients is aimed at checking general fitness for surgery, tide, but it is given orally rather than subcutaneously ensuring that the diabetic management is appropriate, [23]. Although current experience with both of these and that glycaemic control is reasonable. By ‘inap- drugs is limited, neither has a prolonged action and propriate’ management is meant potentially hazardous so should not have any significant implications for drugs such as the potent and/or long-acting sulphony- surgery. lureas glibenclamide and chlorpropamide. Regrettably, a number of older diabetic persons may still be receiv- 25.7 The practical management of ing such preparations, and their treatment may need to surgery in diabetic patients be updated prior to surgery. There are theoretical reasons for avoiding metformin also, if possible (because of lactic acidosis risk), although this is not strictly 25.7.1 The aims of treatment evidence-based [27]. Increasingly, metformin is being The obvious aims of treatment are avoidance of ex- used in combination with insulin both in type 1 and cess mortality and morbidity. As discussed above, with 2 diabetes, and patients should be advised accordingly 354 CH 25 MANAGING SURGERY IN THE ELDERLY DIABETIC PATIENT

on which of their drugs to avoid preoperatively. This Table 25.1 Checklist for preoperative diabetic is also important with the new long-acting metformin assessment preparations. 1. Assess as outpatient and/or admit 2–3 days earlier than Any assessment of glycaemic control should be usual. made using ‘bedside’ blood glucose monitoring with reagent strips, although the potential inaccuracies of 2. Full medical assessment. Chest X-ray and ECG, electrolytes, serum creatinine. such measurement must be borne in mind [28]; the occasional laboratory plasma glucose level should also be 3. Full diabetic assessment. Four times daily bedside blood checked and, if possible, the preoperative glycosylated glucose levels, HbA1c, autonomic function, etc. haemoglobin (HbA1c) level. As mentioned previously, 4. Optimize diabetic management; avoid excessively ‘excellent control’ is not necessary, but significant hy- long-acting hypoglycaemic agents. −1 perglycaemia (e.g. consistently >10.0 mmol l ) will 5. Ensure reasonable glycaemic control.∗ require action to be taken. This may involve moving 6. Liaise closely with the anaesthetist. patients from diet to sulphonylureas, increasing tablet doses if already receiving oral agents, or perhaps intro- ∗See Section 25.7.2 for suggested target blood glucose levels ducing insulin on a temporary basis. If the latter step is required, thrice-daily short-acting insulin (e.g. Ac- 25.7.3 Management in trapid, Humulin S) or short-acting analogues, insulin non-insulin-requiring diabetes lispro (Humalog) or insulin aspart (Novorapid), with or without evening isophane insulin, or twice-daily There is general agreement that diabetic patients not 30/70 premixes (e.g. Humulin M3, Human Mixtard receiving insulin treatment, undergoing surgery of less 30, Novomix 30, Humalog Mix 25) are usually suit- than major severity, can be managed conservatively able. Such patients will of course require preoperative by observation only [2, 32–35]. Surprisingly, there treatment as for ‘insulin -requiring’ diabetic patients. has been very little critical evaluation of this pre- Other preoperative assessments in the elderly should sumed optimal therapy, although the information that is include checking for autonomic neuropathy. This can available does support a conservative approach. Thus, be achieved with simple electrocardiography testing Thompson and colleagues [36] measured plasma glu- (e.g. R-R ratio standing and lying – the ‘30:15 ra- cose and metabolite responses in three groups of male tio’; or during deep breathing) [29]. However, postural patients undergoing transurethral surgery to the bladder hypotension in the elderly diabetic patient does not or prostate gland. The groups were non-diabetic, type always indicate autonomic neuropathy, as it can be 2 diabetic patients treated with intravenous glucose secondary to a wide variety of precipitants, especially and insulin (‘GKI infusion’) [37], and type 2 patients drugs. This is important because autonomic neuropathy treated conservatively. There was no significant dif- (which is more common in the elderly) may occasion- ference between the two diabetic groups in terms of ally be associated with sudden perioperative death [30], perioperative and postoperative blood glucose levels. as well as increased intraoperative morbidity [31]. The The plasma insulin and metabolite levels were actually anaesthetist must be aware of such information, so that closer to the non-diabetic group in the diabetic patients patients can receive close cardiac monitoring. not treated with insulin. The authors concluded that It can be appreciated that much of the standard ‘GKI’ in this situation induced an abnormal metabolic preoperative assessment can be carried out prior to state, with no overall glycaemic benefit. The results of admission, either by liaison with the patient’s physician this study were of general importance, but especially or at a pre-admission anaesthetic clinic. Although this so for those caring for elderly diabetic patients requir- has been advocated for many years [6], regrettably it ing surgery. The patients studied had a mean age of rarely occurs, and patients continue to need admission about 65 years, and type 2 disease – by far the most at least a day or two earlier than usual. Moreover, common type of diabetes among the elderly. surgery may need to be further delayed if unforeseen Guidelines for the conservative management of di- problems are discovered following admission. abetes in surgery for type 2 patients are listed in A summary checklist of the preoperative diabetic as- Table 25.2. It must again be emphasized that ma- sessment is provided in Table 25.1 (note the important jor surgery in such patients (e.g. opening the ab- final step of liaising with the anaesthetist). dominal or thoracic cavity) should be managed as 25.7 THE PRACTICAL MANAGEMENT OF SURGERY IN DIABETIC PATIENTS 355

Table 25.2 Guidelines for surgical care in diabetics not glucose [27, 39]. Not surprisingly, these systems did receiving insulin treatment.∗ not work well! Later systems involved subcutaneous insulin with subsequent intravenous glucose infusions. 1. Operate in the morning if possible. Many such systems were complex, with widely varying 2. Frequently monitor bedside blood glucose levels (e.g. types, amounts and proportions of insulin, and equally 2-hourly). varied concentrations and rates of glucose infusion. 3. If receiving oral agents, omit on morning of surgery, and The results with such methods were variable, although restart with first postoperative meal. in good hands they were comparable with more modern 4. Avoid glucose and lactate-containing fluids intra- methods [40, 41]. venously. Subcutaneous methods of insulin delivery have, however, now been generally abandoned because they 5. Liaise with the anaesthetist. are awkward and inflexible. Continuous intravenous in- ∗Unless surgery is of major severity, or preoperative control sulin and glucose delivery was introduced during the poor – in which case GKI infusion is advisable. late 1970s [42], and became rapidly popular because of its flexibility and simplicity. There are two major for insulin-treated patients. Also of note is the im- methods of use: portance of avoiding glucose-containing intravenous • solutions, which can greatly destabilize glycaemic con- The ‘separate-line’ system (see Figure 25.3). Here, trol. Lactate-containing fluids (e.g. Ringer’s lactate and insulin is infused continuously via a syringe pump, Hartmann’s fluid) should also not be used as they can with glucose infused separately. The glucose infu- have hyperglycaemic effects [38]. Close plasma glu- sion is usually 10% dextrose, delivered at a rate of −1 −1 cose monitoring is of course essential, and again liaison 100 ml h (10 g h ) via an electric drip counter. with the anaesthetist important – it is often in operat- Insulin is usually delivered via a 50 ml syringe ing theatre that the dextrose or Ringer’s drip is erected! driver – 50 units of soluble insulin (e.g. Acrapid, −1 Surgery in the morning is advisable for all types of di- Humulin S) in 50 ml of 0.9% saline. Thus, 1 ml h −1 abetes; there are no special metabolic reasons for this is equivalent to 1 unit h , and an average starting −1 but, from a practical point of view, it is much easier infusion rate is 3 units h . The glucose infusion (and safer) to manage postoperative control problems rate is kept constant throughout, and the insulin rate in the afternoon rather than the middle of the night! varied according to frequent (e.g. 1–2-hourly) bed- These simple management principles are success- side blood glucose measurements, aiming to main- −1 ful in almost all type 2 patients. Very rarely, exces- tain levels in the range of 6–12 mmol l . Whilst it sive postoperative hyperglycaemia may occur, and this can be appreciated that this system is highly flexi- should be managed with subcutaneous, short-acting ble and simple, it is very ‘high-tech’, and requires insulin with meals. Alternatively, if the patient can- expensive equipment which may not always be avail- not tolerate food, a GKI infusion (if the blood glu- able. There is also a potential for ‘metabolic disaster’ cose is very high, an insulin infusion may initially if one of the lines comes adrift. The interruption of be required to bring the glucose levels down to about glucose will lead to dangerous hypoglycaemia, while 10–12 mmol l−1), following which a GKI infusion can the cessation of insulin will conversely lead to hy- be used. perglycaemia and possibly ketosis. • 25.7.4 Management in insulin-requiring The GKI infusion (see Figure 25.3). A simpler and more ‘user-friendly’ version of the above method is diabetes to combine glucose and insulin in the same infusion This group includes true type 1 diabetes patients, type bag, and give them together. A small amount of 2 patients on insulin treatment (including those on a potassium is added to avoid hypokalaemia – hence combination of oral hypoglycaemics and insulin), and the term ‘GKI’ (glucose–KCl–insulin). Interestingly, patients with type 2 diabetes requiring temporary per- this method was first advocated in 1963 by Galloway operative insulin because of poor glycaemic control or and Shuman [7], but did not become popular until planned major surgery. Historically, a confusing num- re-described by Alberti and Thomas in 1979 [39], ber of systems have been advocated at various times; and modified by the same group in 1982 [37]. these include early bizarre systems such as the com- The present most widely used ‘mix’ is 500 ml of plete omission of insulin, or insulin with no subsequent 10% dextrose with 15 units of soluble insulin and 356 CH 25 MANAGING SURGERY IN THE ELDERLY DIABETIC PATIENT

10% Add : 10 mmol KCI dextrose 500 ml 10% and 15 U dextrose soluble insulin (substitute bag with different insulin dosage if 100 mL/h blood glucose levels rise or Drip fall beyond counter Vein acceptable pump limits) Soluble insulin 100 mL/h 50 U in 50 mL saline

Vein

Syringe-driver pump Titrate rate from hourly blood glucose measurement

Figure 25.3 The ‘separate line’ (left) and ‘GKI infusion’ (right) systems of delivering intravenous insulin and glucose to diabetic patients undergoing surgery. (Reproduced by kind permission of Blackwell Scientiﬁc Publications Ltd.) [to be reproduced from 2nd edition].

10 mmol KCl, delivered at 100 ml h−1. This will (whether GKI or ‘separate line’; see Figure 25.3), is provide 3 units of insulin and 10 g glucose per of no metabolic consequence, and it is a matter of hour, as does the ‘separate-line’ technique described practicalities as to which is chosen. Many hospitals above, but without the need for pumps and drip use GKI in the general ward situation, and ‘separate counters. The current (1999) report of the European lines’ in high-dependency or intensive care situations. Diabetes Policy Group [43] suggests that 16 units The GKI system certainly works well in practice of soluble insulin be used in the 10% dextrose bag, and is supported by a study of 85 episodes of and that the infusion is run at 80 ml h−1. No reason surgery using GKI, in which mean plasma glucose is given for the variation but either system will work, levels ranged from 8.3 to 10.2 mmol l−1 during as the differences are inconsequential. Of course if the operative day and ﬁrst two postoperative days the patient has a tendency to hyperkalaemia (e.g. [45]. A summary algorithm for management of the with renal impairment), then the potassium content surgical diabetic patient is shown in Figure 25.4. of the bags must be adjusted, or potassium avoided This includes a suggested scheme for altering GKI altogether. infusions if necessary according to bedside blood Because the insulin and glucose in GKI infusions are glucose monitoring. When the GKI and ‘separate line’ delivered together, the potential metabolic problems of systems were compared in a random fashion [46], the rate alterations, line blockages, and so on, do not exist. separate lines were generally preferred by nursing The main disadvantage of GKI is that if dose changes staff, and resulted in more blood glucose levels within are needed, the whole bag must be discarded and a the target range. The length of hospital stay, duration fresh one prepared and erected. In practice, however, of insulin infusion and numbers of untoward incidents this occurs in only 10–20% of cases [37]. were similar with both systems. Importantly, the study As well as being practically effective, glucose– report did not provide details of the level of nursing insulin infusion systems have advantageous metabolic care available – a major advantage of GKI is that it is effects, reducing the counter-regulatory hormone relatively safe on busy, low-dependency wards. stress response, and improving insulin sensitivity [44]. Electrolytes should be measured daily in patients on The actual method of delivering insulin and glucose GKI infusion, in case hyponatraemia develops. Finally, 25.7 THE PRACTICAL MANAGEMENT OF SURGERY IN DIABETIC PATIENTS 357

Diabetes – Surgery

• Assess • Stabilize • Liaise with anaesthetist

Non-insulin- Insulin-treated Non-insulin-treated treated (other (all patients) (major surgery) surgery)

{PRIVATE}Erect GKI* 8:00–9:00 am Observe BG levels Check BG strips 2-hourly Omit OHAs Aim for BGs of 6–12 mmol/l Restart with first meal • If BG <6 mmol/l, reduce to 10 units GKI • If BG >12 mmol/l, increase to 20 units GKI Check daily BG and U&E Restart s/c insulin when eating

*Standard GKI: 500 ml 10% dextrose + 15 units soluble + 10 mmol KCl. The solution is infused at 100 mlh−1.

Abbreviations: BG = blood glucose; OHA = oral hypoglycaemic agents; U&E = urea and electrolytes; GKI = glucose–potassium–insulin infusion; s/c = subcutaneous.

Figure 25.4 Summary chart for managing diabetes during surgery. it should be noted that both systems described here a continuous carbohydrate provision, 12-hourly subcu- can be used with 5% dextrose if desired, halving taneous isophane insulin, or possibly long-acting ana- the insulin dose delivered as appropriate. Similarly, if logues once or twice daily [48], may be satisfactory. the infused volume needs to be reduced – as may be Potential problems exist with both of these regimens, the case in elderly patients with cardiac problems, or however [47]. With intermittent feeds, once-daily sub- patients with other causes of fluid overload such as cutaneous insulin may be initially introduced, although renal failure – then 20% dextrose at half the volume again twice-daily insulin may be necessary [49]. can be used, or a standard 10% GKI system used at 50 ml h−1 rather than 100 ml h−1 (this would be equivalent to a 5% dextrose system). 25.7.5 Special surgical situations The management of insulin-requiring diabetic pa- Emergency surgery tients being fed preoperatively or postoperatively via a nasogastric tube, or with parenteral nutrition, may Truly urgent surgery is, fortunately, relatively rare in cause some management difficulties. Insulin infusions diabetic patients. However, it does occur and is perhaps are safe if the patient is in a high-dependency or in- more common in the elderly; examples include periph- tensive care area, but in general wards subcutaneous eral and mesenteric embolization, ruptured aneurysms insulin can be used. There is little trial evidence avail- and trauma. Urgent plasma glucose, urea and elec- able examining the various insulin regimens under trolyte estimates are of course mandatory, and if the these circumstances [47]. Provided that the feed has glycaemic and metabolic status is adverse it is best to 358 CH 25 MANAGING SURGERY IN THE ELDERLY DIABETIC PATIENT correct this as far as possible prior to surgery. The ur- During recent years, the use of GKI in diabetic pa- gency of the surgical situation will, of course, need to tients undergoing CABG has resulted in an improve- be assessed carefully in each case, and no ‘blanket’ ment in outcome measures such as length of hospi- rules can be made. Similarly, as regards the method of tal stay, risk of arrhythmias and ventilation duration, peroperative glycaemic control, this too must be de- when compared to those managed without GKI [57]. cided on an individual basis. Generally, a GKI system Recently, a combination of intravenous and then sub- will be advisable in these unplanned situations, but the cutaneous insulin was shown to result in similar levels amounts of insulin needed cannot always be predicted. of morbidity and mortality in diabetic and non-diabetic For example, if the patient requiring urgent surgery has patients following CABG [58]. received a sulphonylurea drug or insulin injection with the past 12 hours, the amount of insulin in the GKI infusion may need to be reduced. Each situation must be 25.8 Conclusions judged individually, and very frequent bedside blood glucose monitoring is necessary. In many cases, it may Although few centres are currently investigating the be judged that a ‘separate line’ system of delivery may practical and theoretical aspects of the operative care be indicated to provide extra flexibility. of the diabetic patient, the subject remains a popular topic for review articles [33, 59]. Perhaps this is not surprising. As a clinical problem it is very common, Open-heart surgery and although current management procedures are well Cardiac surgery is today very common, and coro- accepted, their detailed application continues to cause nary artery bypass grafting (CABG) in particular is confusion. well established as a symptom-relieving and some- Safe and effective perioperative diabetic care re- times life-prolonging operation of low mortality. As quires the acceptance of hospital-based agreed protocols of care which must be widely distributed. Trans- such, it is now being offered to many relatively el- ferring these protocols to safe and effective patient derly patients. Compared with other forms of surgery, treatment, depends on a team approach by physician, CABG is unusual in that it is a long and unusually surgeon and anaesthetist. traumatic operation. Moreover, the patients are also The importance of an appropriate management of in- rendered hypothermic and later given large doses of inpatients with diabetes has been recognized nationally in otropic agents after the restoration of cardiac activity. England by the diabetes National Service Framework All of these factors will promote increased insulin de- (NSF). The NSFs are long-term government strategies mands. Additionally, however, it has been traditional to for improving specific areas of care, and the diabetes use a dextrose ‘priming solution’ to fill the cardiopul- NSF was launched in 2001. The equally important NSF monary pump, often amounting to a glucose load of for older people (published in the same year as the about 75 g at the start of surgery [50]. Not surpris- diabetes NSF) includes standard 4 which covers 16 ingly, the initial results using standard GKI systems core topics in areas such as nutrition, fluid balance in this situation were poor [50, 51], but changing to and pressure area care. Despite these two strategies, a non-glucose priming solution greatly improved the much remains to be done to optimise the perioperative peroperative glycaemic control [51, 52]. A ‘separate care of elderly people with diabetes, in England and line’ system for glucose and insulin provision is essen- elsewhere. tial, with frequent blood glucose monitoring. Relatively large insulin doses are needed [53], but the results are good and algorithms have been produced to aid References insulin delivery decisions [54]. When using modern systems such as these, the results of open-heart surgery 1. Kamoi K. Good long-term quality of life without dia- in diabetic patients have been comparable with those betic complications with 20 years of continuous sub- in non-diabetic counterparts [55, 56]. Hence, the most cutaneous insulin infusion therapy in a brittle diabetic important aims of diabetes management during surgery elderly patient. Diabetes Care 2002; 25: 402–3. have been fulfilled – acceptable glycaemic control by 2. Allison SP, Tomlin PJ and Chamberlain MJ. Some a simple and logical system, without excess mortality effects of anaesthesia and surgery on carbohydrate and or morbidity. fat metabolism. Brit J Anaesth 1979; 41: 588–93. 25.8 CONCLUSIONS 359

3. Elliott MJ and Alberti KGMM. Carbohydrate 19. Bott S, Tusek C, Jacobsen LV et al. Insulin detemir un- metabolism – effects of preoperative starvation and der steady-state conditions: no accumulation and con- trauma. Clin Anaesthesiol 1983; 1: 527–50. stant metabolic effect over time with twice daily ad- 4. Nordenstrom J, Sannenfield J and Arner P. Characteri- ministration in subjects with type 1 diabetes. Diabetic sation of insulin resistance after surgery. Surgery 1989; Medicine 2006; 23: 522–8. 105: 28–35. 20. Mathieu C and Gale EAM. Inhaled insulin: gone with 5. Hume DM and Egdahl RH. The importance of the brain the wind? Diabetologia 2008; 51: 1–5. in the endocrine response to injury Ann Surg 1959; 150: 21. Rosenstock J, Saltes-Rak E, Sugimoto D et al.Triple 697–712. therapy in type 2 diabetes. Diabetes Care 2006; 29: 6. Gill GV and Alberti KGMM. Surgery and diabetes. 554–9. Hospital Update 1989; 15: 327–36. 22. DeFronzo RA, Kim DD, Ratner RE, Fineman MS, 7. Galloway JA and Shuman CR. Diabetes and surgery. Han J and Baron AD. Effects of exanetide (Exendin-4) Am J Med 1963; 34: 177–91. on glycaemic control and weight over 30 weeks in 8. Hjortrup A, Sorenson C, Dynemose E, Hjortso NC and metformin-treated patients with type 2 diabetes. Dia- Kehlet H. Influence of diabetes mellitus an operative betes Care 2005; 28: 1092–100. risk. Brit J Surg 1985; 72: 783–5. 23. Goldstein BJ, Johnson J, Feinglos MN, 9. Melliere D, Berrahal D, Desgranges P et al. Influence Williams-Herman DE and Lunceford JK. Effect of diabetes on revascularisation procedures of the aorta of initial combination therapy with sitagliptin, a and lower limb arteries: early results. Eur J Vasc En- dipeptidyl peptidase-4 inhibitor, and metformin dovasc Surg 1999; 17: 438–41. on glycemic control patients with type 2 diabetes. 10. Hamdan AD, Saltzberg SS, Sheahan M et al. Lack Diabetes Care 2007; 30: 1979–87. of association of diabetes with increased postoperative 24. Van den Berghe G, Wouters P, Weekers F et al. (2001) mortality and cardiac morbidity. Arch Surg 2002; 137: Intensive insulin therapy in critically ill patients. New 417–21. Engl J Med, 345, 1359–67. 11. MacKenzie CR and Charlson ME. Assessment of pe- 25. Coursin DB and Murray MJ. How sweet is euglycaemia rioperative risk in the patient with diabetes mellitus. in critically ill patients? Mayo Clin Proc 2003; 78: Surg Gynaecol Obstet 1988; 167: 293–9. 1460–2. 12. Sandler RS, Maule WF and Baltus ME. Factors as- 26. Watkinson P, Barber VS and Young JD. Strict glucose sociated with post-operative complications in diabetics control in the critically ill. Br Med J 2006; 332: 865–6. after biliary tract surgery. Gastroenterology, 1986; 91: 27. Gill GV and Alberti KGMM. (1992) The care of 157–62. the diabetic patient during surgery. In: International 13. Lieberman D, Friger M and Lieberman D. Rehabilita- Textbook of Diabetes Mellitus. KGMM Alberti, RA tion outcome following hip fracture surgery in elderly deFronzo, H Keen and P Zimmet (eds). John Wiley diabetics: A prospective cohort study of 224 patients. & Sons, Chichester, pp. 1173–83. Disability and Rehabilitation 2007; 29: 339–45. 28. Hutchison ASA and Shenkin A. BM strips: how ac- 14. Lipscombe LL, Jamal SA, Booth GL and Hawker GA. curate are they in general wards? Diabetic Medicine The risk of hip fractures in older individuals with 1984; 1: 225–6. diabetes. Diabetes Care 2007; 30: 835–41. 29. Ewing DJ, Martyn CN, Young RJ and Clarke BF. The 15. Golden SH, Kao WHL, Peart-Vigilance C and Broncati value of cardiovascular autonomic function: 10 years’ FL. Perioperative glycemic control and the risk of experience in diabetes. Diabetes Care 1985; 8: 491–8. infectious complications in a cohort of adults with 30. Page MM and Watkins PJ. Cardiorespiratory arrest with diabetes. Diabetes Care 1999; 22: 1408–14. diabetic autonomic neuropathy. Lancet 1978; i: 14–16. 16. Jennings AM, Wilson RM and Ward JD. Symptomatic 31. Burgos LG, Ebert TJ, Asiddao C, Turner LA, Pattison hypoglycaemia in NIDDM patients treated with oral CZ, Wang-Cheng R and Kampine JP. Increased intra- hypoglycaemic agents. Diabetes Care 1989; 12: operative cardiovascular morbidity in diabetics with au- 203–8. tonomic neuropathy. Anaesthesiology 1989; 70: 591–7. 17. Faull CM, Holmes V and Baylis PH. Water balance in 32. Podolsky S. Management of diabetes in the surgical elderly people: is there a deficiency of vasopression? patient. Med Clin North America 1982; 66: 1361–72. Age and Ageing 1993; 22: 114–20. 33. Hirsch IB, McGill JB, Cryer PE and White PF. Periop- 18. Ashwell SG, Gebbie J and Home PD. Twice-daily erative management of surgical patients with diabetes compared with once-daily insulin glargine in people mellitus. Anaesthesiology 1991; 74: 346–59. with type 1 diabetes using meal-time insulin aspart. 34. Alberti KGMM and Marshall SM. (1988) Diabetes and Diabetic Medicine 2006; 23: 879–86. surgery. In: The Diabetes Annual Ed KGMM Alberti 360 CH 25 MANAGING SURGERY IN THE ELDERLY DIABETIC PATIENT

and LP Krall (eds). Elsevier Publishers, Amsterdam, 48. Putz D and Karabadi UM. Insulin glargine in contin- pp. 248–71. uous enteric tube feeding. Diabetes Care 2002; 25: 35. Schade DS. Surgery and diabetes. Med Clin of North 1889–90. America 1988; 72: 1531–43. 49. Jain N, John LJ, McGlinchey L et al. Audit of the 36. Thompson J, Husband DJ, Thai AC and Alberti management of glycaemia in hospitalised enterally fed KGMM. Metabolic changes in the non-insulin patients with diabetes. Diabetic Med 2008; 25 (Suppl. dependent diabetic undergoing minor surgery: effect 1): 101. of glucose-insulin-potassium infusion. Brit J Surg 50. Gill GV, Sherif IH and Alberti KGMM. Management 1986; 73: 301–4. of diabetes during open heart surgery. Brit J Surg 1981; 37. Alberti KGMM, Gill GV and Elliott MJ. Insulin de- 68: 171–2. livery during surgery in the diabetic patient. Diabetes 51. Stephens JW, Krause AH, Petersen CA et al. The effect Care 1982; 5: 65–77. of glucose priming solutions in patients undergoing 38. Thomas DJB and Alberti KGMM. The hyperglycaemic coronary artery bypass grafting. Ann Thorac Surg 1988; effects of Hartmann’s solution in maturity-onset diabet- 45: 544–7. ics during surgery. Brit J Anaesth 1978; 50: 185–8. 52. Crock PA, Ley CJ, Martin IK et al. Humoral and 39. Alberti KGMM and Thomas DJB. The management metabolic changes during hypothermic coronary artery of diabetes during surgery. Brit J Anaesth 1979; 51: bypass surgery in diabetic and non-diabetic subjects. 693–710. Diabetic Medicine 1988; 5: 47–52. 40. Thomas DJB, Platt HS and Alberti KGMM. 53. Elliott MJ, Gill GV, Home PD et al. A comparison of Insulin-dependent diabetes during the peri-operative two regimens for the management of diabetes during period. Anaesthesia 1984; 39: 629–37. open-heart surgery. Anaesthesiology 1984; 60: 364–8. 41. Gill GV. Surgery and diabetes mellitus. In Textbook of 54. Watson BG, Elliott MJ, Pay DA et al. Diabetes mellitus Diabetes J. Pickup and G. Williams (eds). Blackwell and open heart surgery. A simple practical closed Publishers, London, 1991, pp. 820–6. loop insulin infusion system for blood glucose control. 42. Taitelman U, Reece EA and Bessman AN. Insulin in Anaesthesia 1986; 41: 250–7. the management of the diabetic surgical patient. Con- 55. Lawrie GM, Morris GC and Glaeser DH. Inﬂuence tinuous intravenous administration versus subcutaneous of diabetes mellitus on the results of coronary bypass administration. JAMA 1977; 237: 658–60. surgery. Follow-up of 212 diabetic patients 10 to 15 43. European Diabetes Policy Group 1998. A desktop years after surgery. JAMA 1986; 256: 2967–71. guide to Type 1 (insulin-dependent) diabetes mellitus. 56. Devinen R and McKenzie FN. Surgery for coronary Diabetic Medicine 1999; 16: 253–66. artery disease in patients with diabetes mellitus. Can J 44. Nygren JO, Thorell A, Soop M et al. Perioperative Surg 1985; 28: 367–70. insulin and glucose infusion maintains normal insulin 57. Lazar HL, Chipkin S, Philippides G et al.Glucose- sensitivity after surgery. Amer J Physiol 1998; 275: insulin-potassium solutions improve outcomes in dia- E140–8. betics who have coronary artery operations. Ann Tho- 45. Husband DJ, Thai AC and Alberti KGMM. rac Surg 2000; 70: 145–50. Management of diabetes during surgery with 58. Schmeltz LR, DeSantis AJ, Thiyagarajan V et al.Re- glucose-insulin-potassium infusion. Diabetic Medicine duction of surgical mortality and morbidity in diabetic 1986; 3: 69–74. patients undergoing cardiac surgery with a combined 46. Simmons D, Morton K, Laughton S and Scott DJ. A intravenous and subcutaneous insulin glucose manage- comparison of two intravenous insulin regimens among ment strategy. Diabetes Care 2007; 30: 823–8. surgical patients with insulin-dependent diabetes mel- 59. Anonymous. Drugs in the peri-operative period: corti- litus. Diabetes Educator 1994; 20: 422–7. costeroids and therapy for diabetes mellitus. Drug and 47. Clement S, Braithwaite SS, Magee MF et al.Man- Therapeutics Bulletin 1999; 37: 68–70. agement of diabetes and hyperglycaemia in hospitals. Diabetes Care 2004; 27: 553–91. SECTION V Management of Associated Complications

Edward W. Gregg and Linda Geiss Division of Diabetes Translation, Centers for Disease Control and Prevention, Atlanta, GA, USA

some of the realities of the diabetes epidemic. Namely, Key messages diabetes remains largely a disease of ageing, and this is likely to be even more the case in the future [7–9]. • Diabetes leads to a variety of vascular and Both, the incidence and prevalence of diabetes have in- neuropathic complications which increase the creased substantially in relative terms among the young likelihood of disability. • In older people, because of the increased risk and middle-aged adults. Age is one of most impor- of other geriatric syndromes, diabetes may be tant risk factors for diabetes, roughly equivalent to associated with further disability. the body mass index (BMI) as a continuous, graded • The excess disability risk associated with dia- risk factor. The prevalence of total diabetes (diagnosed betes may lead to a higher prevalence of falls and and undiagnosed diabetes combined) among persons fractures and may underpin the development of aged >65 years in the US was 31% in 2005–2006 [9]. frailty. However, these estimates hide considerable variation across ethnic groups, as more than one-third of older Mexican-Americans and African-Americans have diabetes. Furthermore, prevalence has been estimated at 26.1 Diabetes and older adults: The 25% in studies of nursing homes in both the US and magnitude and character of United Kingdom [10, 11]. the problem The results of multi-national studies conducted during the mid- to late 1990s in Europe suggested that The increase in diabetes prevalence in the United the prevalence of diabetes among Northern European States, western Europe and many areas of the devel- populations (e.g. UK, Netherlands, Poland, Sweden) oping world in recent decades is an ominous threat was only slightly lower than the American estimates, to public health [1–3]. While the threat of increased whereas estimates from Southern European popula- diabetes incidence in young adults and adolescents tions (e.g. Italy, Spain, Malta) were even higher [12]. associated with obesity has garnered much attention In a multi-national study of Asian cohorts, older adults [4–6], these observations may have overshadowed in China and Japan had a total diabetes prevalence

Diabetes in Old Age Third Edition Edited by Alan J. Sinclair © 2009 John Wiley & Sons Ltd. ISBN: 978-0-470-06562-4 364 CH 26 THE IMPLICATIONS OF THE EVOLVING DIABETES EPIDEMIC FOR DISABILITY IN OLDER ADULTS similar to that of the US (i.e. approximately one in [3]. Because these complications are a function of five), whereas prevalence exceeded one in three in diabetes duration and are also influenced by age, older many areas of India [13]. adults with diabetes are likely to have several of these The steep association between age and type 2 dia- conditions simultaneously. betes has practical implications for future projections In addition to the traditional diabetic complications, of diabetes burden in older adults. Because the absolute diabetes increases the risk for several components of prevalence is so high among older adults, the gradual the geriatric syndrome, including cognitive decline increases in prevalence of diabetes observed across the and Alzheimer’s disease, sarcopenia, depression, os- entire age range has a much more substantial absolute teoporotic fractures and incontinence [15, 16]. The end impact among the prevalence of the old than the young. result of this convergence of vascular, neuropathy and For example, among adults age 65–74 years, the preva- geriatric conditions is a complex of functional disabil- lence of diagnosed diabetes has doubled during the ities that are of higher prevalence in people with di- past 25 years, adding nine prevalent cases per 100 to abetes (Figure 26.2). These functional disabilities em- the population (i.e. from 9.1 to 18.5%) (Figure 26.1). anate from impairments in mobility, cognitive decline, Prevalence also increased by 125% among persons depression and other emotional or affective disorders, aged 0–44 years, but this increase added less than one and social isolation. For the older adult with diabetes, case per 100 to the population (i.e. from 0.6 to 1.4%) the implications of these impairments for early loss of (8). As a result, within 25 years, most cases of diagnosed diabetes will be among the population age ≥65 independence and quality of life are as concerning as years, and the total number of cases is expected to triple the more traditionally recognized clinical implications [7]. Accordingly, many of the developed countries of of diabetes [17]. Diabetes may contribute to functional the world will see the most substantial increases in total disabilities through the multiple complications of di- numbers of diabetes cases among older adults [14]. abetes or, alternatively, it may influence functional Diabetes presents a unique problem for older adults, disability through the more direct effects on glycaemic because it leads to a variety of vascular and neuropathic lability, such as through symptom distress associated complications that are likely to be further compounded with hyperglycaemia or hypoglycaemia. by age. Diabetes is the leading cause of blindness, In addition to the obvious implications for the in- non-traumatic amputation and end-stage renal disease, dividual, disability prevalence associated with diabetes and more than doubles the risk of coronary heart is important from the perspective of understanding the disease (CHD), stroke and peripheral arterial disease trends of health status of the population, as well as the

Age (years) 20 18 65–74 16 75+ 14 12 45–64 10 8

Prevalence (%) 6 4 2 0–44 0

1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 Year

National Diabetes Surveillance System, 2008;

Figure 26.1 Prevalence of diagnosed diabetes in the United States, by age, between 1980 and 2006. 26.2 DIABETES AND FUNCTIONAL DISABILITY 365

Functional disability physical cognitive Diabetes Hyperglycaemia social Hypoglycaemia psychological Obesity Sarcopenia Eye disease Renal disease Neuropathy Stroke Coronary heart disease Lower extremity disease

Figure 26.2 Model of potential long-term impact of diabetes on aging-related morbidity. effectiveness of public health interventions in the di- the prevalence of inability to perform key tasks of mo- abetic population [18, 19]. Fries et al. have described bility, including walking 0.4 km (quarter-mile), doing the compression of morbidity as the ultimate objective housework and walking up stairs [25]. For example, of public health efforts for the adult population [19]. 32% of women with diabetes reported being unable to For population morbidity to be compressed, medical walk 0.4 km, compared to about 14% of non-diabetic and public health efforts will have to lead to reduc- women. A similar excess disability existed for climb- tions in disability incidence that are even more rapid ing steps and doing housework. The excess prevalence than mortality reductions. The end result would be a re- of mobility problems extended to objective measures duced average number of years that the average person of physical functioning, including tests of balance, gait spends in a disabled state. speed and lower-extremity function, as reflected in chair stands. Diabetes was associated with a similar In this chapter, the relationship between diabetes risk for mobility impairments in men, although the ab- and functional disability is described, and the promi- solute prevalence of impairments tended to be lower. nent explanatory factors and modifiable risk factors for These findings were corroborated by another nationally functional decline summarized and prioritized. In ad- representative survey, the National Health Interview dition, the most promising modifiable risk factors and Survey, which showed that diabetes was associated interventions are outlined. with about twice the prevalence of inability to perform diverse mobility tasks, with the greatest excess 26.2 Diabetes and functional risk observed for lower-extremity tasks [26]. These disability cross-sectional studies have the important caveat of not being able to determine whether diabetes preceded Several epidemiological studies have examined the as- disability, or vice versa. Nevertheless, they played an important role in defining the extent of morbidity in sociation of diabetes with functional disabilities, with the population with diagnosed diabetes. most focusing on physical impairments. During the Prospective studies further support the association 1990s, diabetes was shown to be an important predictor between diabetes and the development of disability, of maintenance of independence and successful aging demonstrating a 100–150% increased incidence of dis- in diverse cohorts, including adults up to the age of 75 ability among older women with diabetes relative to years in the Alameda County Study, the Framingham their non-diabetic peers [27]. In the Study of Osteo- Study, and the US Longitudinal Study of Aging and porotic Fractures (SOF), highly functioning community the Established Population for the Epidemiologic Study dwelling women with diabetes developed an inability of the Elderly Survey [20–24]. The steep rise in dia- to walk 0.4 km mile at the rate of 4% per year, an in- betes prevalence during the 1990s in turn led to more ability to do housework at a rate of 8% per year, and detailed and comprehensive examinations of the par- 10% per year developed an inability to do at least one ticular role of diabetes on physical disability. In an ex- task (Figure 26.3) [27]. For example, the 5-year cu- amination of the US non-institutionalized population, mulative incidence of disability – defined by onset of women and men with diabetes consistently had twice inability to perform major physical tasks – was about 366 CH 26 THE IMPLICATIONS OF THE EVOLVING DIABETES EPIDEMIC FOR DISABILITY IN OLDER ADULTS

12 No Diabetes 10 Diabetes

4 Yearly incidence (%)

0 Walking1/4 Heavy Climbing10 Prepare Shopping Any Task mile housework steps meals

Figure 26.3 Yearly incidence of inability to do physical and household tasks among women aged ≥65 years, with and without diabetes. Data taken from the Study of Osteoporotic Fractures [28].

40% for 65–80-year old women with diabetes, com- that occur before the onset of diabetes, such as obesity, pared to about 20% in their same age non-diabetic insulin resistance, inflammation, CHD and peripheral counterparts. In the SOF, having diabetes was equiv- arterial disease. Other disabilities that follow may re- alent to about 7 years of aging in terms of functional sult from diabetes, such as peripheral neuropathy, di- status. The Women’s Health and Aging Study (WHAS) abetic nephropathy, vision loss, cardiovascular disease found similar findings in a cohort of older women with and stroke, while other factors are directly associated mild to moderate impairments. Women with diabetes with glycaemic status (e.g. hyperglycaemia symptoms, had a 78% increased risk of mobility-related disability hypoglycaemia). Among the US population, CHD was and a 65% increased risk of ADL disability [37]. a dominant factor, explaining about one-third of the ex- The results of several studies have suggested that the cess disability in diabetic women and about one-quarter excess disability risk associated with diabetes extends in diabetic men. The BMI also explained a large por- to an increased risk of falls and osteoporotic fractures. tion of disability, but only in women (24%), as the In NHANES III, diabetic women had a 45% higher BMI and CHD combined explained 52% of the excess prevalence of falls (35% in the past year) compared to disability [25]. In men, the BMI was relatively unim- non-diabetic women (25%), and a 69% increased risk portant but was offset somewhat by stroke, which ex- of an injurious fall in the previous year. In the WHAS, plained about one-fifth of the excess disability. About women with diabetes had a 50% increased risk of falls. one-third of the excess disability risk remained unex- The increased frailty and risk of falls is sufficient to plained by all complications included in the statistical increase hip fracture risk even though people with type analyses. The observation that BMI was a more im- 2 diabetes tend to have a higher bone mineral density portant intervening factor in women than in men was (BMD) than their non-diabetic peers [28–30]. People also consistent with a study of national sample of older with type 1 diabetes, on the other hand, have both a community-dwelling adults (diabetic and non-diabetic decreased BMD and an increased risk of falls, creating combined) [33]. Men with class III obesity (BMI >40) a double jeopardy in terms of fracture risk [31, 32]. had an 83% increased disability risk, but there was no excess disability risk associated with overweight or moderate obesity (i.e. BMI 30–35). In women, there 26.3 Modifiable factors explaining were 15%, 54% and 83% increases in the risk of diabetes and disability disability associated with class I, II and III obesity, respectively. Diabetes could lead to physical disability through a va- Subsequent studies with a better assessment riety of pathophysiological pathways, including some of lower-extremity conditions showed that the 26.4 INTERVENTIONS TO REDUCE DISABILITY RISK 367 ankle:brachial index and sensory and motor nerve of both atherosclerotic and neuropathic origin, appear function each had strong correlations with objective to be among the most dominant factors because they measures of mobility and balance, and were potent exact a large magnitude of increased disability risk. predictors of incident disability [34–36]. Analyses Yet, at the same time they are fairly common among the of the Womens Health and Aging Study (a cohort population with diabetes, with lower-extremity disease with moderate physical impairment at baseline) (either peripheral arterial disease or peripheral neu- found that peripheral arterial disease, peripheral ropathy) being present in 38% of adults aged 70–80 neuropathy and depression were all important years, in 45% of those aged >80 years, and likely predictors, each accounting for up to one-fifth of the in an even higher proportion of the subset with dia- excess disability in older diabetic women [37]. Again, betes [41]. Here, CHD plays a similar role, at least a large proportion of excess mobility and activities doubling the incidence of disability and affecting per- of daily living (ADL) disability, and about 40% of haps one-quarter of the older diabetic population. Other excess walking limitations and performance measures, factors such as stroke and hip fracture have profound remained unexplained by all factors combined. effects on disability when they occur, but fortunately These findings were further supported by subclinical are less common than lower-extremity conditions on assessments of functioning, defined by objective the whole and thus may act as secondary factors. Fur- physical performance tests such as walking speed, ther research is required into the intervening factors in balance and chair stands [38]. A follow-up of the disability, and how interventions can be developed to Health ABC Study showed that older adults with type exploit them. 2 diabetes have a greater decline in leg muscle mass, strength and muscle quality (strength per unit of mass) than those without diabetes. Of note, this difference 26.4 Interventions to reduce in the rate of decline was limited to lower-extremity disability risk strength, which the authors speculated may be due to neuropathy, which affects the lower extremities more Observational studies have pointed to a wide range of dramatically than upper extremities [39]. potentially effective interventions to reduce the risk of Whereas, the studies described above examined fac- disability in diabetic populations, but have rarely been tors explaining the difference in disability risk be- tested using randomized clinical trials. The strongest tween diabetic and non-diabetic population, other stud- evidence exists for exercise, strength and balance train- ies have examined the predictors of poor physical ing among persons with moderate impairments. Studies function and disability specifically within the diabetic conducted during the 1990s showed that older adults population. In a sample of over 1000 managed-care at a high risk of falls can improve their physical per- patients with diabetes, obesity, lack of exercise, depres- formance and reduce their risk of falling when ran- sion symptoms and lower formal education were each domized to structured programmes of lower-extremity associated with disability [40]. In the SOF, physical ac- strength, balance training, walking and tai chi [42–44]. tivity and obesity were each prominent modifiable risk These programmes have typically been aimed at people factors for incident disability [27]. Obese women de- who already have an established functional impair- veloped an inability to perform physical tasks at twice ment. More recent studies of overweight and obese the rate of lean women, while women in the highest adults with arthritis symptoms have shown that struc- quintile of physical activity had a 39% lower disabil- tured exercise programmes also reduce symptoms, im- ity risk compared to sedentary women. In a detailed prove mobility and maintain independence [45]. The examination of the association of diabetes with risk of fact that exercise also improves functional capacity in falls among diabetic older women, those who used in- persons with lower-extremity arterial disease, and is sulin and had a glycated haemoglobin (HbA1c) level of also an aide to weight maintenance, are further jus- <6% had a fourfold higher risk of falls than women tification to recommend structured, moderate exercise with higher HbA1c levels [40]. Peripheral nerve dys- programmes. While such studies have not been con- function, poor vision, poor renal function and weight ducted specifically among adults with diabetes, suitable loss were each associated with a 40–60% increased risk evidence for collateral benefits on glycaemic control of falls. without adverse events exists to justify the general- In summary, the route between diabetes and physical ization of exercise programmes to older adults with disability is multifactorial. Lower-extremity diseases, diabetes [46]. Accordingly, the American College of 368 CH 26 THE IMPLICATIONS OF THE EVOLVING DIABETES EPIDEMIC FOR DISABILITY IN OLDER ADULTS Sports Medicine and the American Heart Association adults are a positive step in the reduction of disability have revised their recommendations for physical activ- [15]. ity, suggesting that older adults should have an activity plan that includes appropriate levels of regular aerobic, muscle-strengthening, flexibility and balance exercise, 26.5 Trends in disability among the and also integrates preventive and therapeutic recom- diabetic population mendations specific to individual chronic conditions present with the individual patient [47]. Population trends in the prevalence and incidence of Depression is common, underdiagnosed and under- disability over time will help to evaluate the cumula- treated in adults with diabetes [48, 49]. The fact that tive effects of public health efforts to reduce morbidity depression was a factor explaining excess disability in among the diabetic population. Meta-analyses of US diabetic patients suggests that better identification and national surveys have suggested that the prevalence of treatment could also reduce disability [27, 37]. Im- disability has declined consistently among adults aged proved glycaemic control has been shown to reduce >65 years during the past 20 years [55]. Such decline symptom distress and improve physical functioning in disability prevalence has been attributed to a com- over the short term. However, despite the clear benefits bination of wide-scale improvements in medical care, of glycaemic control in terms of reduced microvascu- key health behaviours such as smoking, the availability lar complications, the impact on long-term disability of assistive devices and general improvements in the remains unclear. In addition, the recent controversial socioeconomic status of the population [56]. However, findings of the ACCORD study, in which intensively whether these improvements have also occurred among managed older adults had a higher risk of death than the US population with diabetes remains unclear. those under simply good control, will likely lead to Data from the US National Diabetes Surveillance a further review of the optimal levels of glycaemic System suggest that there have been modest reductions control among men [58]. Structured weight-loss proin disability prevalence over the past decade. Declines grammes have been suggested as a means of reducing in the prevalence of mobility limitation among the dia- disability in high-risk populations, based on observa- betes population have been greatest among black men tions of the impact of obesity on disability risk com- (i.e. a decline from 60% to 38%) and Hispanic men bined with the collateral benefits of weight loss in (52% to 36%) [57] (Figure 26.4). These findings are terms of exercise capacity and reduced cardiovascular risk. On the other hand, the potential for loss of consistent with reduced rates of hospitalization associ- lean mass associated with weight loss could conceiv- ated with complications among the diabetic population ably negate the advantages of fat loss. Observational [57]. However, part of the decline in disability preva- studies have associated weight loss with either no ben- lence may have been driven by an increased incidence efit or increased disability and mortality, pointing to of diabetes diagnosis and an accompanying lower av- the need for randomized controlled trials [50–52]. The erage duration of diabetes in the overall diabetic popu- ongoing Look AHEAD study, which ostensibly was lation. Disability prevalence has also declined in black designed to examine the impact of intensive long-term and Hispanic women by 7 percentage points, but has weight loss in obese diabetic adults, should also pro- been flat in white women [57]. vide important information on the impact of weight loss on disability [53, 54]. Given the role of CHD, stroke and lower-extremity 26.6 Summary disease on the development of disability over the long term, cardiovascular risk factor management including Disability is a central complication of diabetes in older smoking cessation, lipid and lipid and blood pressure adults, and represents the cumulative impact of the nu- control, aspirin use and early screening for diabetic merous chronic, acute and transient complications of complications, may have long-term effects on disabil- diabetes. Since the prevalence of diabetes is also heav- ity. However, such interventions may be far upstream ily influenced by demographic and mortality trends from effects on disability, and both clinical trial and in the population, an understanding of the trends in observational studies have not (yet) demonstrated any morbidity and disability will require novel and more effects on disability. Nevertheless, recent attempts to sensitive approaches to measuring disability among the tailor guidelines and implications to the needs of older population. The reduction in disability among people 26.6 SUMMARY 369

60 White Black Females 50 Hispanic White 40 Black Males Hispanic 30

Prevalence (%) 10

0 1997 1998 1999 2000 2001 2002 2003 2004 2005 Year

Figure 26.4 Trends in age-adjusted prevalence of any mobility limitation per 100 adults with diabetes, by race/ethnicity and gender in the United States, 1997–2005. Data available from the US national diabetes surveillance system. with diabetes will continue to depend on the devel- 8. Centers for Disease Control and Prevention. Preva- opment and implementation of new interventions at lence of Diagnosed Diabetes by Age, United States, individual, health system and health policy levels. 1980–2000. http://www.cdc.gov/diabetes/statistics/ prev/national/fig2 2003. Available from: URL: www. cdc.gov/diabetes/statistics/prev/national/fig2 9. Cowie CC, Rust KF, Ford ES, Eberhardt MS, References Byrd-Holt DD et al. Full accounting of diabetes and pre-diabetes in the U.S. population in 1988–1994 and 1. Wild S, Roglic G, Green A, Sicree R and King H. 2005–2006. Diabetes Care 2009; 32: 287–294. Global prevalence of diabetes: estimates for the year 10. Resnick HE, Heineman J, Stone R and Shorr RI. 2000 and projections for 2030. Diabetes Care 2004; 27 Diabetes in U.S. nursing homes, 2004. Diabetes Care (5): 1047–53. 2008; 31 (2): 287–8. 2. Zimmet P, Alberti KG and Shaw J. Global and societal 11. Sinclair AJ, Gadsby R, Penfold S, Croxson SC and implications of the diabetes epidemic. Nature 2001; 414 Bayer AJ. Prevalence of diabetes in care home resi- (6865): 782–7. dents. Diabetes Care 2001; 24 (6): 1066–8. 3. Engelgau MM, Geiss LS, Saaddine JB, Boyle JP, 12. Age- and sex-specific prevalences of diabetes and im- Benjamin SM, Gregg EW et al. The evolving diabetes paired glucose regulation in 13 European cohorts. Di- burden in the United States. Ann Intern Med 2004; 140 abetes Care 2003; 26 (1): 61–9. (11): 945–50. 4. Dabelea D, Pettitt DJ, Jones KL and Arslanian SA. 13. Qiao Q, Hu G, Tuomilehto J, Nakagami T, Balkau B, Type 2 diabetes mellitus in minority children and Borch-Johnsen K et al. Age- and sex-specific preva- adolescents. An emerging problem. Endocrinol Metab lence of diabetes and impaired glucose regulation in 11 Clin North Am 1999; 28 (4): 709–29, viii. Asian cohorts. Diabetes Care 2003; 26 (6): 1770–80. 5. Olshansky SJ, Passaro DJ, Hershow RC, Layden J, 14. King H, Aubert RE and Herman WH. Global burden of Carnes BA, Brody J et al. A potential decline in life diabetes, 1995–2025: prevalence, numerical estimates, expectancy in the United States in the 21st century. N and projections. Diabetes Care 1998; 21 (9): 1414–31. Engl J Med 2005; 352 (11): 1138–45. 15. Brown AF, Mangione CM, Saliba D and Sarkisian CA. 6. Williams DE, Cadwell BL, Cheng YJ, Cowie CC, Guidelines for improving the care of the older person Gregg EW, Geiss LS et al. Prevalence of impaired with diabetes mellitus. J Am Geriatr Soc 2003; 51 (5 fasting glucose and its relationship with cardiovascular Suppl. Guidelines): S265–80. disease risk factors in US adolescents, 1999-2000. 16. Gregg EW and Brown A. Cognitive and physical dis- Pediatrics 2005; 116 (5): 1122–6. abilities and aging-related complications of diabetes. 7. Boyle JP, Honeycutt AA, Narayan KM, Hoerger TJ, Clinical Diabetes 2003; 21: 113–18. Geiss LS, Chen H et al. Projection of diabetes burden 17. Phelan EA, Anderson LA, La Croix AZ and Larson through 2050: impact of changing demography and EB. Older adults’ views of ‘successful aging’ – how disease prevalence in the US. Diabetes Care 2001; 24 do they compare with researchers’ definitions? J Am (11): 1936–40. Geriatr Soc 2004; 52 (2): 211–16. 370 CH 26 THE IMPLICATIONS OF THE EVOLVING DIABETES EPIDEMIC FOR DISABILITY IN OLDER ADULTS

18. Fried LP, Tangen CM, Walston J, Newman AB, Hirsch 32. Strotmeyer ES and Cauley JA. Diabetes mellitus, bone C, Gottdiener J et al. Frailty in older adults: evidence mineral density, and fracture risk. Curr Opin En- for a phenotype. J Gerontol A Biol Sci Med Sci 2001; docrinol Diabetes Obes 2007; 14 (6): 429–35. 56 (3): M146–56. 33. Imai K, Gregg EW, Chen YJ, Zhang P, de RN and 19. Fries JF. Measuring and monitoring success in com- Williamson DF. The association of BMI with func- pressing morbidity. Ann Intern Med 2003; 139(5Pt 2): tional status and self-rated health in US adults. Obesity 455–9. (Silver Spring) 2008; 16 (2): 402–8. 20. Black SA, Ray LA and Markides KS. The prevalence 34. Dolan NC, Liu K, Criqui MH, Greenland P, Gural- and health burden of self-reported diabetes in older nik JM, Chan C et al. Peripheral artery disease, dia- Mexican Americans: ﬁndings from the Hispanic es- betes, and reduced lower extremity functioning. Dia- tablished populations for epidemiologic studies of the betes Care 2002; 25 (1): 113–20. elderly. Am J Public Health 1999; 89 (4): 546–52. 35. Resnick HE, Stansberry KB, Harris TB, Tirivedi M, 21. Boult C, Kane RL, Louis TA, Boult L and McCaffrey Smith K, Morgan P et al. Diabetes, peripheral neu- D. Chronic conditions that lead to functional limitation ropathy, and old age disability. Muscle Nerve 2002; in the elderly. J Gerontol 1994; 49 (1): M28–36. 25 (1): 43–50. 22. Guccione AA, Felson DT, Anderson JJ, Anthony JM, 36. Brach JS, Solomon C, Naydeck BL, Sutton-Tyrrell K, Zhang Y, Wilson PW et al. The effects of speciﬁc Enright PL, Jenny NS et al. (2008) Incident physical medical conditions on the functional limitations of disability in people with lower extremity peripheral elders in the Framingham Study. Am J Public Health arterial disease: the role of cardiovascular disease. J 1994; 84 (3): 351–8. Am Geriatr Soc, 56 (6), 1037–44. 23. Moritz DJ, Ostfeld AM, Blazer D, Curb D, Taylor JO 37. Volpato S, Blaum C, Resnick H, Ferrucci L, Fried and Wallace RB. The health burden of diabetes for the LP and Guralnik JM. Comorbidities and impairments elderly in four communities. Public Health Rep 1994; explaining the association between diabetes and lower 109 (6): 782–90. extremity disability: The Women’s Health and Aging 24. Strawbridge WJ, Cohen RD, Shema SJ and Kaplan GA. Study. Diabetes Care 2002; 25 (4): 678–83. Successful aging: predictors and associated activities. 38. de Rekeneire N, Resnick HE, Schwartz AV, Shorr RI, Am J Epidemiol 1996; 144 (2): 135–41. Kuller LH, Simonsick EM et al. Diabetes is associated 25. Gregg EW, Beckles GL, Williamson DF, Leveille SG, with subclinical functional limitation in nondisabled Langlois JA, Engelgau MM et al. Diabetes and physical older individuals: the Health, Aging, and Body Com- disability among older U.S. adults. Diabetes Care 2000; position study. Diabetes Care 2003; 26 (12): 3257–63. 23 (9): 1272–7. 39. Park SW, Goodpaster BH, Strotmeyer ES, Kuller LH, 26. Ryerson B, Tierney EF, Thompson TJ, Engelgau MM, Broudeau R, Kammerer C et al. Accelerated loss of Wang J, Gregg EW et al. Excess physical limitations skeletal muscle strength in older adults with Type among adults with diabetes in the U.S. population, 2 diabetes: the health, aging, and body composition 1997–1999. Diabetes Care 2003; 26 (1): 206–10. study. Diabetes Care 2007; 30 (6): 1507–12. 27. Gregg EW, Mangione CM, Cauley JA, Thompson TJ, 40. Schwartz AV, Vittinghoff E, Sellmeyer DE, Feingold Schwartz AV, Ensrud KE et al. Diabetes and incidence KR, de RN, Strotmeyer ES et al. Diabetes-related of functional disability in older women. Diabetes Care complications, glycemic control, and falls in older 2002; 25 (1): 61–7. adults. Diabetes Care 2008; 31 (3): 391–6. 28. Lipscombe LL, Jamal SA, Booth GL and Hawker 41. Gregg EW, Sorlie P, Paulose-Ram R, Gu Q, Eberhardt GA. The risk of hip fractures in older individuals MS, Wolz M et al. Prevalence of lower-extremity with diabetes: a population-based study. Diabetes Care disease in the US adult population>=40 years of age 2007; 30 (4): 835–41. with and without diabetes: 1999-2000 national health 29. Robbins J, Aragaki AK, Kooperberg C, Watts N, and nutrition examination survey. Diabetes Care 2004; Wactawski-Wende J, Jackson RD et al. Factors associ- 27 (7): 1591–7. ated with 5-year risk of hip fracture in postmenopausal 42. Campbell AJ, Robertson MC, Gardner MM, Norton women. JAMA 2007; 298 (20): 2389–98. RN, Tilyard MW and Buchner DM. Randomised con- 30. Schwartz AV, Nevitt MC, Brown BW, Jr and Kelsey trolled trial of a general practice programme of home JL. Increased falling as a risk factor for fracture among based exercise to prevent falls in elderly women. Br older women: the study of osteoporotic fractures. Am Med J 1997; 315 (7115): 1065–9. J Epidemiol 2005; 161 (2): 180–5. 43. Gregg EW, Pereira MA and Caspersen CJ. Physical 31. Ivers RQ, Cumming RG, Mitchell P and Peduto AJ. activity, falls, and fractures among older adults: a Diabetes and risk of fracture: The Blue Mountains Eye review of the epidemiologic evidence. J Am Geriatr Study. Diabetes Care 2001; 24 (7): 1198–203. Soc 2000; 48 (8): 883–93. 26.6 SUMMARY 371

44. Province MA, Hadley EC, Hornbrook MC, Lipsitz LA, 51. Lee JS, Kritchevsky SB, Tylavsky FA, Harris T, Ever- Miller JP, Mulrow CD et al. The effects of exercise on hart J, Simonsick EM et al. Weight-loss intention in the falls in elderly patients. A preplanned meta-analysis well-functioning, community-dwelling elderly: associ- of the FICSIT Trials. Frailty and Injuries: Cooperative ations with diet quality, physical activity, and weight Studies of Intervention Techniques. JAMA 1995; 273 change. Am J Clin Nutr 2004; 80 (2): 466–74. (17): 1341–7. 52. Lee JS, Kritchevsky SB, Tylavsky F, Harris T, Si- 45. Messier SP, Loeser RF, Miller GD, Morgan TM, monsick EM, Rubin SM et al. Weight change, weight Rejeski WJ, Sevick MA et al. Exercise and dietary change intention, and the incidence of mobility lim- weight loss in overweight and obese older adults itation in well-functioning community-dwelling older with knee osteoarthritis: the Arthritis, Diet, and Ac- adults. J Gerontol A Biol Sci Med Sci 2005; 60 (8): tivity Promotion Trial. Arthritis Rheum 2004; 50 (5): 1007–12. 1501–10. 53. Rejeski WJ, Lang W, Neiberg RH, Van DB, Foster 46. Boule NG, Haddad E, Kenny GP, Wells GA and Sigal GD, Maciejewski ML et al. Correlates of health-related RJ. Effects of exercise on glycemic control and body quality of life in overweight and obese adults with mass in Type 2 diabetes mellitus: a meta-analysis Type 2 diabetes. Obesity (Silver Spring) 2006; 14 (5): of controlled clinical trials. JAMA 2001; 286 (10): 870–83. 1218–27. 54. Wadden TA, West DS, Delahanty L, Jakicic J, Rejeski 47. Nelson ME, Rejeski WJ, Blair SN, Duncan PW, J, Williamson D et al. The Look AHEAD study: a Judge JO, King AC et al. Physical activity and pub- description of the lifestyle intervention and the evi- lic health in older adults: recommendation from the dence supporting it. Obesity (Silver Spring) 2006; 14 American College of Sports Medicine and the Amer- (5): 737–52. ican Heart Association. Circulation 2007; 116 (9): 55. Manton KG, Corder L and Stallard E. Chronic dis- 1094–105. ability trends in elderly United States populations: 48. Anderson RJ, Freedland KE, Clouse RE and Lustman 1982–1994. Proc Natl Acad Sci USA 1997; 94 (6): PJ. The prevalence of comorbid depression in adults 2593–8. with diabetes: a meta-analysis. Diabetes Care 2001; 24 56. Cutler DM. The reduction in disability among the el- (6): 1069–78. derly. Proc Natl Acad Sci USA 2001; 98 (12): 6546–7. 49. Dealberto MJ, Seeman T, Mcavay GJ and Berkman L. 57. Centers for Disease Control and Prevention. Diabetes Factors related to current and subsequent psychotropic Surveillance System: Health Status and Disability. Dia- drug use in an elderly cohort. J Clin Epidemiol 1997; betes Surveillance System 2004. Available from: URL: 50 (3): 357–64. http://www.cdc.gov/diabetes/statistics/health status 50. Launer LJ, Harris T, Rumpel C and Madans J. Body national.htm. mass index, weight change, and risk of mobility dis- 58. Gerstein HC, Miller ME, Byington RP et al. Action to ability in middle-aged and older women. The epidemi- Control Cardiovascular Risk in Diabetes Study Group. ologic follow-up study of NHANES I. JAMA 1994; The Effects of Intensive Glucose Lowering in Type 2 271 (14): 1093–8. Diabetes. N Engl J Med 2008; 358: 2545–59. 27 Diabetes and Cognitive Dysfunction

Alan J. Sinclair1 and Koula G. Asimakopoulou2 1Bedfordshire and Hertfordshire Postgraduate Medical School, University of Bedfordshire, Luton, UK 2King’s College London Dental Institute, Oral Health Services Research and Dental Public Health, London, UK

The idea that diabetes may be related to cognitive Key messages dysfunction makes intuitive sense. The brain relies on a steady supply of glucose in order to function op- • Diabetes mellitus and cognitive dysfunction are timally, but at the same time is unable to store glu- likely to be interrelated, and often coexist in the cose in situations of over-supply (e.g. hyperglycaemia). same individual. Equally, it cannot rely on stored facilities in situa- • There is an increasing evidence base linking tions of under-supply (hypoglycaemia). It follows that a greater likelihood of cognitive dysfunction a condition associated with abnormalities in glucose in subjects with diabetes, especially of long availability to the brain, such as diabetes, might lead duration. • The assessment of cognitive function using stan- to brain processes becoming compromised. dard cognitive screening tests is recommended Cognitive function is a short descriptive term for the in the routine assessment of all older people with multitude of mental processes that people engage in diabetes. in everyday life, from paying attention to and perceiv- • The detection of cognitive dysfunction may pro- ing information, to remembering it, organizing it into vide an early opportunity to consider drug-based meaningful chunks, retrieving it and using it to solve intervention strategies and care packages that abstract or practical problems. Cognitive psychologists provide more effective management, and may refer to such processes as ‘cognitive domains’, and delay the need for early dependency. classify them – among others – in terms of attention, speed and amount of information processing (mental ﬂexibility), and memory. A trauma or insult to the brain 27.1 Introduction such as that posed by hyper- or hypoglycaemia, can affect any or all of these processes selectively. Older Since the early 1920s, when Miles and Root [1] ﬁrst patients with diabetes (either type 1 or 2) have an in- proposed a link between diabetes and cognitive dys- creased risk of developing cognitive dysfunction by function, knowledge concerning the relationship be- virtue of their increasing age, irrespective of other fac- tween these two states – as well as the methods of tors such as diabetes itself. Practising clinicians must investigation of such a relationship – have progressed expect to see patients with both diabetes and cognitive enormously. dysfunction, as both conditions are highly prevalent

Table 27.1 Background to the relationship between measures used [5]. Such studies have varied not only diabetes and cognitive disorders [2]. in the sensitivity, specificity, reliability and validity of the tests they have employed, but also in the degree of • Professional and public concern about the impact of diabetes on cognition methodological vigour they have adopted, with the tendency of studies revealing more diabetes-related cogni- • Long-term influence of hyperglycaemia and hypogly- tive deficits the less well-controlled they have been [6]. caemia on cerebral function unknown Nevertheless, the overall consensus based on a review • Pathophysiological mechanisms involved uncertain, but of the vast majority of such studies seems to be that: may involve both vascular, inflammatory and neuronal mechanisms “...patients with type 2 diabetes have moderate impairments across all cognitive domains ...a dimin- • No current agreement on the most optimum method to ished ability to efficiently process unstructured infor- detect/assess cognitive deficits in diabetes mation, particularly when the cognitive task at hand • Clinical relevance of the changes observed uncertain. requires speed of response.” [7] Given the shortcoming of cross-sectional investigations, however, and the reported consensus that such chronic disease states in today’s’ communities. The de- studies are probably ill-equipped [3, 7, 8] to answer the velopment of dysfunction may have several important question as to whether diabetes patients are cognitively consequences for the patient and his/her family and impaired, the remainder of this section relies on the carers in terms of the complexity of care, adherence findings of reviews (e.g. [9]) and systematic reviews to therapy, ability to self-manage, and the need for (e.g. [10]) of longitudinal studies in demystifying the assisted care. Many of the important elements which relationship between diabetes and cognition. have provided a background to this area are listed in In one of the first prospective studies of cognitive Table 27.1. decline in diabetes [11], the investigators examined an all-female North American sample of women as part of 27.2 Evidence of association a wider study on osteoporotic fractures. Using a mod- between diabetes and ified Mini-Mental State Examination (MMSE; [12], measuring dementia), Digit Symbol Substitution (DSS; cognitive dysfunction [13] a measure of psychomotor speed) and Trail Mak- ing B (TMB; [14]) measuring sustained visual attention In a recent editorial published in the British Medical and mental shift, it was reported that women with di- Journal [3], it was argued that abetes were twice more likely than their diabetes-free “...diabetes, cognitive impairment and dementia are counterparts to show major cognitive decline. strongly linked, but the precise mechanisms are un- A year later, Fontbonne et al. [15] reported from a clear.” French sample as part of the Epidemiology of Vascular Aging Study. Having classified the study participants This apt conclusion describes the findings of a, now in terms of glycaemic profile at baseline [normal fast- substantial, body of literature on the subject spanning ing blood glucose, impaired (6.1–6.9 mmol l−1) fasting 15 years or so, and which has attempted to iden- glucose and diabetes], they were followed up for 4 tify, what (if any) cognitive processes are mostly af- years and then assessed for a sizeable battery of cog- fected by diabetes, in what ways, and the factors that nitive tests. These tests included the TMB and DSS, as seem to be moderating the diabetes–cognitive dysfunc- well as measures of mental flexibility and auditory at- tion relationship. It is likely that further research will tention (Paced Auditory Serial Addition Test, PASAT; demonstrate that both vascular dementia (VaD) and [16]), psychomotor speed (Finger Tapping Test; [17]), Alzheimer’s disease (AD) are possible outcomes of verbal memory (Auditory Verbal Learning Test, AVLT; long-duration diabetes in a vulnerable individual [4]. [18]), visual memory (Benton Visual Retention Test, The vast majority of studies in this field have been BVRT; [19]) and facial recognition (Facial Recognition cross-sectional, often comparing a small sample of di- Test, FRT; [20]). The authors reported a greater than abetes patients with diabetes-free individuals, on a bat- twofold rate of cognitive impairment in people with di- tery of various cognitive tests, and reporting moderate abetes, having controlled for confounding factors such differences in some – but usually not all – cognitive as age, education and gender. 27.2 EVIDENCE OF ASSOCIATION BETWEEN DIABETES AND COGNITIVE DYSFUNCTION 375

In a similar set-up, Kanaya et al. [21] reported 4-year • a greater rate of decline in cognitive function. follow-up data from a subset of participants in the Ran- Interestingly, the authors noted that their results had cho Bernardo Study, a prospective trial which had been probably underestimated the deleterious effects of di- running in California since 1972. Both men and women abetes on cognitive function, and cited two reasons. (n = 999) were classified in terms of glycaemic status First, the reviewed studies tended to exclude people (normal, impaired glucose tolerance and diabetes) and who already had some form of cognitive impairment assessed on three cognitive tests, namely the MMSE, at baseline, and as such, selectively sampled ‘healthier’ TMB and a verbal fluency test assessing semantic individuals with a lower subsequent risk of cognitive memory (VF). The authors reported no age-adjusted decline. Second, most of the data that they reviewed differences in baseline cognitive function scores as a failed to include information on people who died or function of glycaemic group. Four years later, however, were lost at follow-up; it is argued that success in being the women with diabetes had a fourfold increased risk followed-up may in itself be a result of better cognitive of major cognitive decline, as evidenced in impaired function and, as such, discounting people who could verbal fluency test scores. not be followed-up may have simply masked the true At about the same time, a study was conducted to rate of cognitive decline in people with diabetes. assess the risk of cognitive dysfunction and develop- So, how might diabetes be related to cognitive ment of AD in a prospective study of aging and AD decline? In type 1 diabetes, the amount and extent in 824 older (age >55 years) Catholic nuns, priests of exposure to hypoglycaemia have been argued and brothers (this was a subsample of the Religious to be predictive of cognitive decline [25], although Orders Prospective Study) [22]. The participants were a meta-analysis examining the effects of type 1 followed-up for approximately 5.5 years, and had their diabetes on cognition failed to find evidence for this cognitive performance assessed by a robust collection association [26]. of global and specific tests. These included the MMSE, In type 2 diabetes the picture is much more com- DSS, Logical Memory A (assessing ability to remem- plicated; patients with the illness tend to be older ber logical sequences [13]), items from the National and present with other comorbidities such as hyper- Adult Reading Test (NART; a measure of pre-morbid tension, atherosclerotic vascular disease, obesity and intelligence and verbal fluency [23]), Digits Forward depression – which in themselves are independent risk and Backward (assessing working memory and men- factors for cognitive dysfunction. In addition, distal control [13]), items from the Standard Progressive ease duration, glycaemic control, socioeconomic status, Matrices (assessing visuospatial ability [24]), and many age, gender, microvascular complications, insulin re- others. In appropriately adjusted analyses, the authors sistance and the presence of ApoE ε4 allele, may all reported that, not only did the diabetes group have moderate the relationship between diabetes and cogni- lower cognitive function scores at baseline in most of tive dysfunction (for helpful and detailed reviews on the cognitive domains assessed, but that diabetes was these, the reader is referred to Refs [7, 27, 28]). It is also associated with a more rapid (by ca. 44%) rate likely that the aetiology is multifactorial in origin, with of cognitive decline in perceptual speed and a 65% varying contributions from repeated hypoglycaemia, increase in the risk of developing AD. long-duration hyperglycaemia, amyloid deposition, in- The trials reviewed above represent a small subset sulin resistance, cerebrovascular disease, changes in of several studies evaluating the link between diabetes the hypothalamo-pituitary axis and inflammatory dis- and cognitive dysfunction, and are indicative of the ease. overall pattern of results reported in the majority of Another important factor which may exacerbate the published studies in the field. The same results have influence of diabetes on cognitive performance is blood been echoed in a systematic review of prospective pressure control. Data from both the Framingham study studies aiming to evaluate the extent to which diabetes [29] and the OCTO-Twin Study from Sweden [30] is associated with cognitive decline and dementia [10]. have demonstrated that, in patients with type 2 dia- This systematic review concluded that, compared to betes, cognitive performance is worse in the presence diabetes-free individuals, people with diabetes have: of hypertension or an increase in blood pressure. The Framingham cohort included patients aged 55–89 years • a 1.5-fold greater risk of cognitive decline; who, over a long duration, demonstrated poorer results • a 1.6-fold greater risk of developing dementia; and in logical memory scores and word fluency after an 376 CH 27 DIABETES AND COGNITIVE DYSFUNCTION

Verbal fluency scores at • VF Scores adjusted for age, baseline and 4y f/u in women education, depression, 19 lipids, blood pressure, HbA1c, and vascular 18 disease 17 • Significant correlation with 16 HbA1c and baseline (r= − 0.08; p= 0.08) and follow-up VF score 15 (r = −0.10; p = 0.04) VF tests 14 in women only 13 • NGT, normal glucose NGT IGT DM tolerance; IGT, impaired glucose tolerance; DM, Baseline test Follow-up visit diabetes mellitus

Figure 27.1 Association of verbal memory with glucose status: the rancho bernardo study. DM = diabetes mellitus; IGT = impaired glucose tolerance; NGT = normal glucose tolerance; VF = verbal ﬂuency. Reproduced with permission from Ref. [21].

independent increase in blood pressure of 10 mmHg. examination, reduced verbal fluency) and in an in- In the OCTO-Twin Swedish population-based study, creased risk of developing a dementing syndrome [32], sequential MMSE scores over 8 years were signifi- although this has not been a consistent finding [33, cantly lower in the cohort of patients with both diabetes 34]. In the Rancho Bernardo study, which included and hypertension than in the cohorts of either condi- 999 Caucasian subjects aged 42–89 years with varying tion alone, or in those without either condition. This degrees of glucose tolerance who were followed over relationship between cognitive decline and with the 4 years (see Figure 27.1), a significant correlation with presence of either diabetes and hypertension was also HbA1c level and baseline and follow-up verbal fluency observed in the Atherosclerosis Risk in Communities scores were observed, but in women only [21]. (ARIC) study [31] in a 6-year follow-up of almost 11 Various studies have demonstrated a relationship 000 individuals aged 47–70 years at the initial assess- between measures of glycaemia and performance in ment. cognitive assessment. The Stanford (USA) Studies The fact that people with diabetes are at an in- in patients with type 2 diabetes showed that cogni- creased risk of cognitive dysfunction is now active deficits involving verbal learning and complex cepted unequivocally. However, the precise direction perceptual-motor domains were worse in those with of this relationship – as well as its major constituent poorer glycaemic control, and that treatment with a parts – remains unexplored and hence undetermined. sulphonylurea for 7 months led to significant im- Is it the case that diabetes causes cognitive decline in- provements in metabolic control and tests of learning dependently of the moderating factors noted above? and memory [35, 36]. In another study, diabetic pa- Could cognitive decline predispose to developing di- tients aged ≥70 years presenting to a geriatric diabetes abetes? Could it be that a combination of some (or clinic were screened for cognitive dysfunction with all) of the comorbidities noted above may cause dia- the MMSE and a clock-drawing test (CDT) [37]. The betes per se, cognitive decline per se, or perhaps both? CDT scores were inversely correlated with HbA1c lev- Or, finally, is the presence of other comorbidities a els, which suggested that cognitive dysfunction was moderator/mediator of a diabetes–cognitive dysfunc- associated with poor glycaemic control (r =−0.38, tion relationship? The jury is still out. p < 0.004). The relationship between postprandial hyperglycaemia (PPG) and cognitive performance was recently 27.3 Cognitive dysfunction and studied in two groups of older patients with diabetes glycaemic control who were treated with either rapaglinide or glibenclamide [38]. The coefficient of variation of PPG There is evidence that impaired glucose tolerance was found to be associated with MMSE scores (r = (IGT) is associated with poorer performance in cer- −0.3410; p < 0.001) and a composite score of ex- tain cognitive function scores (lower Mini-Mental State ecutive and attention functioning (r =−0.3744; p < 27.4 THE IMPORTANCE OF DETECTING COGNITIVE DYSFUNCTION 377 0.001) after adjusting for multiple confounders. The Impaired cognitive function may result in poorer results suggested that a tighter control of PPG might adherence to treatment, worsen glycaemic control due influence the degree of cognitive decline in older pa- to erratic taking of diet and medication, and increase tients with diabetes. the risk of hypoglycaemia if the patient forgets that In a study of 1983 postmenopausal women (mean he or she has taken the hypoglycaemic medication and age 67 years) the association between HbA1C level repeats the dose. and risk of developing cognitive impairment was determined [39]. Mild cognitive impairment (MCI) or 27.4.1 Methods of detection dementia was seen subsequently to develop over a Cognitive dysfunction has been traditionally assessed 4-year period. For each 1% increase in HbA1C level the women showed a greater age-adjusted likelihood of de- through cognitive tests, using many different proce- veloping MCI (OR = 1.50; 95% CI: 1.14–1.97) and of dures ranging from single global estimates of cognitive developing dementia (OR = 1.40; 95% CI: 1.08–1.83). functioning (e.g. the MMSE) to substantial batteries of neuropsychological assessments spanning the major For those in whom the HbA1C level was ≥7%, the age-adjusted risk for developing MCI was increased al- cognitive domains of language, perception, attention, most fourfold (OR = 3.70; 95% CI: 1.51–9.09) and for memory, visuoconstruction ability, speed of informa- developing dementia was increased almost threefold tion processing and executive (complex) functioning (OR = 2.86; 95% CI: 1.17–6.98). These results clearly [20]. Some of these domains have been less well exam- suggest that, in older patients with IGT or diabetes, the ined than others, and some tests – such as the MMSE levels of glycaemia and cognitive status are linked. and DSS – appear to have been used extensively when comparing people with diabetes with healthy controls. Whilst it is beyond the purpose of this chapter to de- 27.4 The importance of detecting scribe the myriad of tests currently available to assess cognitive dysfunction cognitive performance in older adults, the reader is referred to Lezak’s [20] Neuropsychological Assessment Several benefits may be acquired from the early recog- for a compendium of hundreds of such tests. nition of cognitive impairment in older people with What is of interest here is a distinction that needs to diabetes (see Table 27.2), and this places emphasis be drawn in terms of ‘why’ health care professionals on the importance of tests of cognition as part of might wish to cognitively assess people with diabetes. the functional assessment of older patients. Depend- Until now, only those studies which assess cognitive ing on its severity, cognitive dysfunction in older function by comparing the performance of people with diabetic subjects may have considerable implications diabetes to that of diabetes-free controls have been con- which include increased hospitalization, less ability for sidered. However, by adopting a case-control method self-care, less likelihood of specialist follow-up, and an of assessment, conclusions can be drawn regarding the increased risk of institutionalization [40]. extent to which diabetes patients are in some way impaired in the respective test domain, by comparing Table 27.2 Benefits of early recognition of cognitive mean numerical scores in the diabetes sample with dysfunction in diabetes. those in controls. In the absence of any established test ‘norms’, such a process allows the assessment of • Prompts the clinician to consider the presence of cerebrovascular disease and to review other vascular risk cognitive performance in order to compare functioning factors. between a diabetes group and an appropriately matched control sample. So the answer to ‘why assess’ here is • May be an early indicator of Alzheimer’s disease and to establish cross-group differences. provides early access to medication. Informative though such comparisons may be, they • Allows patients and families to benefit early with social are not particularly useful in a clinical setting, where and financial planning and access to information about it is impractical to obtain appropriate control groups support groups and counselling. or indeed standardized norms for batteries of cognitive • Creates opportunities to consider interventions for tests. Here, clinicians might be assessing cognition to diabetes-related cognitive impairment: optimizing glu- determine whether their patient is currently cognitively cose control; controlling blood pressure and lipids. compromised but, given the test setting, will not have access to a control group. In such cases, the 378 CH 27 DIABETES AND COGNITIVE DYSFUNCTION ‘why’ behind the cognitive assessment is to establish down to 0 for an inability to make any reasonable whether further testing, referral or increased future representation of a clock (for details, see Ref. [44]). cognitive monitoring might be necessary. In these Nishiwaki et al. [45] have shown that, in isolation, cases, it can be argued that a significant proportion of the MMSE might not detect MCI, whilst the CDT cognitive tests which discuss the literature on diabetes might produce a large number of false positives. When and cognitive functioning renders itself beyond use, as used together, however, these tests can be reliable pre- such tests rely on comparing the patient’s performance dictors of moderate to severe cognitive decline. Given with that of others. that their administration and scoring make minimal de- There are two notable exceptions to this general mands in terms of time and resources, it has been observation – the MMSE and the CDT. The MMSE, argued [46] that, from a clinical point of view, they as the most widely used dementia screening, takes only are ideal cognitive functioning screening tests for older 5–10 min to administer, and consists of questions relat- people with diabetes. ing to attention, orientation, memory, calculation and language. It has been criticised as being heavily re- 27.4.2 Influence on diabetes self-care liant on language (and as such may not be suitable for non-English speakers), but it is available in different Although many investigations have been conducted on languages. Typical tasks on the MMSE involve pa- ways to improve the patient’s diabetes self-care and tients being asked to recall the year, month, date, day on their cognitive function, the relationship between and time, and to spell the word ‘world’ backwards. the two states remains under-researched. The question They are also asked to name three objects that are in here is whether cognitive functioning in diabetes pa- the examination room and, a few minutes later, un- tients predicts their efforts to self-manage the illness, expectedly to recall them. Although the MMSE is a with the implication that perhaps a poorer cognitive reliable indicator of moderate to severe cognitive im- performance may be related to poorer self-management pairment, it is not sensitive enough to detect MCI. This skills. may not necessarily be an issue, however, as MCI is In one of the few studies in this area, [40] an not thought to be related to diabetes self-care activities investigation was conducted as to whether cognitive in any significant way [41]. The MMSE is scored out of impairment was associated with changes in self-care 30, with higher scores being indicative of better cog- behaviour and the use of health and social services nitive performance. Specifically, scores of 27–30 are in a community-based case control study of older pa- regarded normal, whiles scores <26 indicate various tients with diabetes. Cognitive function was assessed degrees of cognitive impairment. using the MMSE and CDT, while self-care was as- The CDT is another popular measure that is quick sessed in two ways: (i) by counting the number of pa- and easy to administer. Participants are given a circle tients that were solely responsible for self-medication (no bigger than 4–10 cm in diameter) and are told that and blood glucose (BG) monitoring; and (ii) by moni- it represents a clock face. They are then instructed toring their attendance at a specialist diabetes clinic. to put in the numbers so that the circle now looks Performance on the CDT showed that respectively, like a clock and, when they have done so, to ‘set the 65% and 72% of diabetes patients placed the num- time’ to 10 minutes past 11. The test assesses executive bers and hands correctly, compared to 76% and 84% function and, in particular, the patient’s ability to plan of controls. Age was found to interact with cognitive ahead, their visuospatial ability, ability to engage in dysfunction and self-management, in that older dia- abstract reasoning and, of course, their concentration. betes patients were found to have worse cognitive test The CDT can be scored in several ways ranging in performance, a higher dependency and poorer diabetes amount of detail and precision. An extensive discussion self-management. on CDT administration and scoring is provided by In another study [41], 51 people with type 2 diabetes Shulman et al. [42]. Of the several scoring methods completed a battery of cognitive tests and the Sum- proposed, four-point [43] or five-point [44] systems mary of Diabetes Self-Care Activities questionnaire are probably the quickest and easiest. For example, [47]; however, only a few associations between cogni- when using a five-point scoring system, the patient’s tive functioning and self-management were observed. drawing is assessed from being perfect (score 5), to This lack of association might have been due to the showing inaccurate representation of 10 past 11 when limited statistical power of the study for detecting rela- the overall visuospatial organisation is good (score 3), tionships, or to the absence of any significant practical 27.4 THE IMPORTANCE OF DETECTING COGNITIVE DYSFUNCTION 379 association between self-reported self-care and specific cognitive performance may play a role in medication cognitive skills. One of the few significant associations taking, it fails to explain a substantial amount of pa- that were found was the inverse relationship between tients’ variability in this behaviour. self-reported memory problems (as assessed by the More recently, Trimble et al. [52] assessed the abil- Subjective Memory Questionnaire [48]) and the num- ity of the CDT to predict problematic insulin ad- ber of diabetes problem-solving strategies (as assessed ministration skills in older adults with diabetes. A by Toobert and Glasgow’s Diabetes Problem-Solving group of 30 patients who had not used insulin before Interview), although self-reported memory complaints were taught to self-administer a sham insulin injec- were not a reliable indicator of objective cognitive tion with an insulin pen, using a standardized protocol. function. A better dietary self-management was pre- The injections were continued for 7 days, after which dicted by a better general (Modified Wisconsin Card self-administration was re-tested. An abnormal CDT Sorting Test [49]) and diabetes-specific abstract rea- was significantly associated with more problems in soning. Better exercise self-management was predicted learning to perform the sham injections (measured as by better scores on a test of mental flexibility, the those who were unable to correctly complete all steps Serial Subtractions of 7s [20], and generating more of the protocol, or those who omitted all or part of diabetes-specific problem-solving strategies was pre- a step), although a small number of patients with a dicted by fewer subjective memory problems. The re- normal CDT also demonstrated major problems. The searchers assessed self-reported self-care through the results were in line with those of other studies, which Summary of Diabetes Self Care Activities (SDSCA noted the frequency of abnormal CDTs in older pa- [47]). In a later study, however, Asimakopoulou and tients [40], and the frequency of errors in older peo- Hampson [50] showed that the SDSCA may be prone ple self-administering insulin [53]; the suggested was to recall biases in people with diabetes and, as such, it made that “...the CDT is a valuable predictor of pois suggested that self-reported self-care as assessed by tential problems with insulin administration skills in instruments such as the SDSCA should be confirmed elderly patients”. by clinical interview and opinion. Finally, Munshi et al. [54] assessed the relationship Other groups have examined self-care on the ba- between global cognitive function as measured by the sis of medication adherence and glycaemic control. MMSE, CDT and Clock in Box (CIB [55]) tests, as For example, Rosen et al. [51] assessed the associ- well as glycaemic control (measured by HbA ) in ation between cognitive performance and adherence 1C older adults with diabetes. Some 34% of patients had to oral hypoglycaemic medication, HbA1c level and missed appointments. Cognitive function was assessed low scores on the CIB, and 38% had low scores on the among other measures with the MMSE, TMB and CDT. Both, the CIB and CDT were superior at iden- the Stroop test (this provides a measure of atten- tifying patients with cognitive dysfunction, compared tion and mental flexibility, with patients being asked to the MMSE. The CIB test was more sensitive in pre- to read out the ink colour of words spelling out dicting poor glycaemic control than the CDT; however, incongruent colour words [20]). Adherence to met- both clock tests were inversely correlated with HbA1C formin was measured using pill bottle caps which levels, which suggested that cognitive function might contained a microprocessor that recorded the date and play a role in the control of diabetes. times of bottle openings; the caps were placed on It appears that the few studies which have assessed the patients’ prescribed antihyperglycaemic medica- the relationship between diabetes self-management and tion. Age was the best predictor of medication ad- cognition have argued for a relationship between cog- herence, and accounted for just under 10% of the nitive dysfunction and impaired self-care, in patients variance in this behaviour. Medication adherence was with diabetes. The amount of variance in self-care be- also predicted by performance on the Stroop word haviours that cognitive tests seem to predict seems test and with TMB completion time, where a worse rather low, however, and additional patient-centred cognitive performance predicted poorer medication research is required in order to elucidate the rela- taking, although the amount of variance explained tionship between cognitive dysfunction and diabetes was only small (<10%). Interestingly, neuropsycho- self-management behaviours. This is particular true logical performance was not associated with HbA1C with regards to the extent to which modest differences levels, but a poor MMSE score predicted missed in cognitive testing might predict practical diabetes appointments. These results suggested that, although self-care skills. 380 CH 27 DIABETES AND COGNITIVE DYSFUNCTION 27.5 The importance of excluding appropriate treatment, or a referral to other specialist depression services.

The presence of a depressive illness may influence the 27.6 Further investigations outcome of any cognitive assessment in older patients with diabetes. Cognitive performance scores are likely The decision to investigate patients with diabetes who to be diminished and create difficulties of interpretation have observed deficits in cognition needs to be taken for the clinician. Since it can be a chronic disorder on the basis of history, examination, impact (if any) with frequent relapse, patients often have difficulty of deficits or behaviour, personality, normal social in maintaining a stable level of glycaemia, and the and professional functioning, and ability for diabetes consequent burden on caregivers can be increased self-care management. substantially. Whilst a full neuropsychological battery of tests will Diabetes appears to be significantly associated with prove helpful, it is not essential in everyday clinical depression, independent of age, gender or the pres- practice. Special techniques such as visual and so- ence of chronic disease [56], while the presence of matosensory evoked potentials are too sophisticated in diabetes appears to double the odds of developing de- routine care, as is electroencephalography. Techniques pression [57]. In one study, the finding of depression such as magnetic resonance imaging (MRI), functional had important implications for a group of in-patients MRI, single photon emission computed tomography as it was the single most important indicator of subse- and offer exciting opportunities to equate cognitive quent death [58]. Failure to recognize depression can performance with definitive evidence of structural and be serious, since it is a long-term, life-threatening, dis- physiological functioning [28, 32]. abling illness and can have a significant impact on the A scheme for the routine screening and detection of patient’s quality of life [59]. Depression may be asso- cognitive dysfunction is shown in Figure 27.2, and this ciated with worsening diabetic control and decreased should serve as a basis for other centres to develop this treatment compliance [60]. approach. It is important that, at the initial assessment, patients undergo a thorough history and examination and in particular are asked about any symptoms of depression. 27.7 Conclusions They should then undergo a mood screening test such as the four-item Geriatric Depression Score [61] or In view of the high prevalence of both diabetes and even shorter instruments. If a significant mood disorder dementing syndromes in aged subjects, every physi- is detected, the opportunity presents itself to offer cian/clinician involved in providing diabetes care to

All diabetic patients aged ≥55 years

Annual review (routine) or Memory disorder or behaviour change

•Informed consent •Brief history •MMSE •Verbal Fluency test •Trails B Test Clock Drawing Test

Low probability of Higher probability of cognitive cognitive impairment impairment

•Exclude depression •Visual assessment

Review annually •Exclude hypoglycaemia Further neuropsychological evaluation; Full history and physical examination; Laboratory evaluation; Imaging

Figure 27.2 Scheme for the detection of cognitive dysfunction in type 2 diabetes mellitus. 27.7 CONCLUSIONS 381 this group should be familiar with this association of prospective observational studies. Diabetologia 48 and be skilled in the initial assessment of cognitive (12): 2460–9. performance. Although the pathogenesis of cognitive 11. Gregg, E.W., Langlois, J.A., Beckels, G.L.A., Engel- dysfunction remains unclear, it can be regarded as a gau, M.M., Williamson, D.F., Narayan, K.M.V. and complication of long-duration diabetes and is likely to Leveille, S.G. (2000) Is diabetes associated with cog- have an important vascular basis. nitive impairment and cognitive decline among older women? Study of Osteoporotic Fractures Research Older subjects may be particularly prone in view Group. Archives of Internal Medicine 160 (2): 174. of other comorbidities, which makes it essential that 12. Folstein MF, Folstein, SE and McHugh PR (1975) a cognitive assessment should form part of an initial Mini-Mental State: A practical method for grading the assessment of a newly diagnosed patient with diabetes, state of patients for the clinician. Journal of Psychiatric and also part of any routine, annual review. Research, 12: 189–98. In this respect, a familiarity with common screen- 13. Weschler, D. (1997) Weschler Adult Intelligence Scale ing methods of cognitive function would be of great - III. The Psychological Corporation, San Antonio, help – as would the recognition that these individuals Texas. require greater specialist care, not less! [62]. 14. Reitan, R.M. and Wolfson, D. (1993) The Halstead-Reitan neuropsychological test battery: Theory and clinical interpretation. 2nd edition. Neuropsychology Press, South Tucson, AZ. References 15. Fontbonne, A., Ducimetiere,` P., Berr, C. and Alper- ovitch, A. (2001) Changes in cognitive abilities over a 1. Miles, W.R. and Root, H.F. 1922. Psychologic tests ap- 4-year period are unfavorably affected in elderly dia- plied to diabetic patients. Archives of Internal Medicine, betic subjects: results of the Epidemiology of Vascular 30, 767–77. Aging Study. Diabetes Care 24 (2): 366. 2. Sinclair AJ 2004. European Diabetes Working Party for 16. Gronwall, D.M.A. (1977) Paced auditory Older People 2001–2004. Clinical Guidelines for Type serial-addition task: A measure of recovery 2 Diabetes Mellitus. Available on: www.eugms.org. from concussion. Perceptual and Motor Skills 44, 3. Strachan, M.W., Price, J.F. and Frier, B.M. (2008) 367–73. Diabetes, cognitive impairment, and dementia. Br Med 17. Mitrushina, M.N., Boone, K.B. and D’Elia, L.F. (1999) J 336 (7634), 6. Handbook of Normative Data for Neuropsychological 4. Biessels GJ, Staekenborg S, Brunner E, Brayne C and Assessment. Oxford University Press, New York. Scheltens P. (2006) Risk of dementia in diabetes – a 18. Spreen, O. and Strauss, E. (1991) A compendium of systematic review. Lancet Neurology, 5: 64–74. neuropsychological tests: Administration, norms and 5. Cosway, R., Strachan, M.W. et al. (2001) Cognitive commentary. Oxford University Press, New York. function and information processing in type 2 diabetes. 19. Benton AL. (1963) The Revised Visual Retention Test, Diabet Med 18 (10): 803–10. 3rd edition. The Psychological Corporation, New York, 6. Asimakopoulou, K.G., S.E. Hampson, et al. (2002) pp. 1–74. Neuropsychological functioning in older people with 20. Lezak, M.D. (1995) Neuropsychological Assessment. type 2 diabetes: the effect of controlling for confound- Oxford University Press, New York. ing factors. Diabet Med 19 (4): 311–16. 21. Kanaya AM, Barrett-Connor E, Gildengorin G and 7. van den Berg, E., R.P. Kessels, et al. (2007) Type 2 Yaffe K (2004) Change in cognitive function by glu- diabetes, cognitive function and dementia: vascular and cose tolerance status in older adults: a 4-year prospec- metabolic determinants. Timely Top Med Cardiovasc tive study of the Rancho Bernardo study cohort. Arch Dis 11: E7. Intern Med 164, 1060–5. 8. Strachan, M.W., I.J. Deary, et al. (1997) Is type II 22. Arvanitakis, Z., R.S. Wilson, et al. (2004) Diabetes diabetes associated with an increased risk of cognitive mellitus and risk of Alzheimer disease and decline in dysfunction? A critical review of published studies. cognitive function. Arch Neurol 61 (5): 661–6. Diabetes Care 20 (3): 438–45. 23. Nelson, H.E. and Willison, J. (1991) National Adult 9. Allen, K.V., B.M. Frier, et al. (2004) The relationship Reading Test Manual. 2nd edition. NFER-Nelson, between type 2 diabetes and cognitive dysfunction: lon- Windsor. gitudinal studies and their methodological limitations. 24. J. Raven, J.C. Raven and J.H. Court (1998) Man- European Journal of Pharmacology 490 (1-3): 169–75. ual for Raven’s progressive matrices and vocabulary 10. Cukierman, T., H.C. Gerstein, et al. (2005) Cognitive scales. Section 1: General overview. Oxford Psycholo- decline and dementia in diabetes–systematic overview gists Press, Oxford. 382 CH 27 DIABETES AND COGNITIVE DYSFUNCTION

25. Deary, I.J. and Frier, B.M. (1996) Severe hypogly- 38. Abbatecola AM, Rizzo MR, Barbieri M, Grella R, caemia and cognitive impairment in diabetes. Br Med Arciello A, Laieta MT, Acampora R, Passariello N, J, 313 (7060): 767–8. Cacciapuoti F and Paolisso G. (2006) Postprandial 26. Brands, A.M., G.J. Biessels, et al. (2005) The ef- plasma glucose excursion s and cognitive functioning fects of type 1 diabetes on cognitive performance: a in aged type 2 diabetics. Neurology 67: 235–40. meta-analysis. Diabetes Care 28(3): 726–35. 39. Yaffe K, Blackwell T, Whitmer RA, Krueger K and 27. Ryan, C.M. (2006) Diabetes and brain damage: more Barrett-Connor E. (2006) Glycosylated haemoglobin (or less) than meets the eye? Diabetologia 49 (10): level and development of mild cognitive impairment 2229–33. or dementia in older women. J Nutr Health Aging 10, 28. Kodl CT and Seaquist ER. (2008) Cognitive dysfunc- 293–5. tion and diabetes mellitus. Endocrine Reviews 29 (4): 40. Sinclair AJ, Girling AJ and Bayer AJ. (2000) Cognitive 494–511. dysfunction in older subjects with diabetes mellitus: 29. Elias PK, Elias MF, D’Agostino RB, Cupples LA, Wil- impact on diabetes self-management and use of care son PW, Silbershatz H and Wolf PA. (1997) NIDDM services. All Wales Research into Elderly (AWARE) and blood pressure as risk factors for poor cognitive Study. Diabetes Res Clin Pract 50 (3): 203–12. performance: The Framingham Study. Diabetes Care 41. Asimakopoulou, K. and Hampson, S.E. (2002) Cogni- 20: 1388–95. tive functioning and self-management in older people 30. Hassing LB, Hofer SM, Nilsson SE, Berg S, Pedersen with diabetes. Diabetes Spectrum 15(2): 116. NL, McClearn G and Johansson B. (2004) Comorbid 42. Shulman, K.I. (2000). Clock drawing: Is it the ideal type 2 diabetes mellitus and hypertension exacerbates cognitive screening test?, International Journal of Geri- cognitive decline: evidence from a longitudinal study. atric Psychiatry 15: 548–61. Age Ageing, 33 (4), 355–61. 43. Death, J., Douglas, A. and Kenny, R.A. (1993) Com- 31. Knopman D, Boland LL, Mosley T, Howard G, Liao D, parison of clock drawing with Mini Mental State Exam- Szklo M, McGovern P, Folsom AR for the Atheroscle- ination as a screening test in elderly acute hospital ad- rosis Risk in Communities (ARIC) Study Investigators. missions. Postgraduate Medical Journal 69: 696–700. (2001) Cardiovascular risk factors and cognitive de- 44. Shulman K.I., Gold, D.P., Cohen, C.A. and Zucchero, cline in middle-aged adults. Neurology, 56 (1), 42–8. C.A. (1993) Clock drawing and dementia in the com- 32. Kodl, C.T. and E.R. Seaquist (2008) Cognitive dys- munity: A longitudinal study. International Journal of function and diabetes mellitus. Endocrine Reviews 29 Geriatric Psychiatry 8: 487–96. (4), 494–511. 45. Nishiwaki, Y., Breeze, E., Smeeth, L., Bulpitt, C.J., 33. Scott RD, Kritz-Silverstein D, Barrett-Commor Peters, R. and Fletcher, AE. (2004) Validity of the E and Wiederholt WC (1998) The association of non-insulin-dependent diabetes mellitus and cognitive Clock-Drawing test as a screening tool for cognitive function in an older cohort. J Am Geriatr Soc 46, impairment in the elderly. American Journal of Epi- 1217–22. demiology 160 (8): 797–807. 34. Lindeman RD, Romero LJ, LaRue A, Yau CL, Schade 46. Asimakopoulou K, Tomlin A and Sinclair A (2008) DS, Koehler KM, Baumgartner RN and Garry PJ. Cognitive function and self-care in type 2 diabetes. (2001) A biethnic community survey of cognition in Diabetes and Primary Care 10 (2): 70–82. participants with type 2 diabetes, impaired glucose tol- 47. Toobert, D.J. and Glasgow, R.E. (1994) Assessing erance, and normal glucose tolerance: the New Mexico diabetes self-management: the Summary of Diabetes Elder Health Survey. Diabetes Care 24, 1567–72. Self-Care Activities questionnaire. In: C. Bradley (ed.), 35. Reaven GM, Thompson LW, Nahum D and Haskins Handbook of Psychology and Diabetes. Harwood Aca- E. (1990) Relationship between hyperglycaemia and demic, Chur, Switzerland. cognitive function in older NIDDM patients. Diabetes 48. Bennett-Levy J and Powell GE (1980) The subjective Care 13, 16–21. memory questionnaire (SMQ). An investigation into 36. Gradman TJ, Laws A, Thompson LW and Reaven the self-reporting of ‘real-life’ memory skills. Br J GM. (1993) Verbal learning and/or memory improves Social and Clinical Psychology, 19, 177–88. with glycemic control in older subjects with 49. Hart RP, Kwentus JA, Wade JB and Taylor JR. (1988) non-insulin-dependent diabetes mellitus. J Am Geriatr Modiﬁed Wisconsin Sorting Test in elderly normal, de- Soc, 41 (12): 1305–12. pressed and demented patients. Clin Neuropsychologist 37. Munshi M, Grande L, Hayes M, Ayres D, Suhl E, 2: 49–56. Capelson R, Lins S, Milberg W and Weinger K. (2006) 50. Asimakopoulou, K. and Hampson, S.E. (2005) Biases Cognitive dysfunction is associated with poor diabetes in self reports of self care in type 2 diabetes. Psychol- control in older adults. Diabetes Care 29, 1794–9. ogy Health and Medicine 11, 305–15. 27.7 CONCLUSIONS 383

51. Rosen, M.I., Beavais, J.E., Rigsby, M.O., et al. (2003) The Osservatorio Geriatrico of Campania Region Neuropsychological correlates of suboptimal adherence Group. Diabetes Metab 22 (5): 314–18. to metformin. Journal of Behavioral Medicine 26 (4): 57. Anderson RJ, Freedland KE, Clouse RE and Lustman 349–60. PJ. (2001) The prevalence of comorbid depression in 52. Trimble, L.A., Sundberg, S., Markham, L., Janicijevic, adults with diabetes: a meta-analysis. Diabetes Care 24 S., Beattie, B.L. and Meneilly, G.S. (2005) Value of the (6): 1069–78. Clock Drawing Test to predict problems with insulin 58. Rosenthal MJ, Fajardo M, Gilmore S, Morley JE and skills in older adults. Canadian Journal of Diabetes 29 Naliboff BD. (1998) Hospitalization and mortality of (2): 102–4. diabetes in older adults. A 3-year prospective study. 53. Coscelli, C., Calabrese, G., Fedele, D., Pisu, E., Diabetes Care 21 (2): 231–5. Calderini, C., Bistoni, S., Lapolla, A., Mauri, M.G., 59. Egede LE, Zheng D and Simpson K. (2002) Comorbid Rossi, A. and Zappella, A. (1992) Use of premixed depression is associated with increased health care use insulin among the elderly. Reduction of errors in and expenditures in individuals with diabetes. Diabetes patient preparation of mixtures. Diabetes Care 15 Care 25 (3): 464–70. (11): 1628–30. 60. Lustman PJ, Anderson RJ, Freedland KE, de Groot M, 54. Munshi, M., Capelson, R., Grande, L., et al. (2006) Carney RM and Clouse RE. (2000) Depression and Cognitive dysfunction is associated with poor diabetes poor glycemic control: a meta-analytic review of the control in older adults. Diabetes Care 29 (8): 1794–9. literature. Diabetes Care 23 (7): 934–42. 55. Grande, L., Milberg, W., Rodolph, J., Gaziano, M. 61. Burke WJ, Roccaforte WH and Wengel SP. (1991) and McGlinchey, R. (2005) A timely screening for The short form of the Geriatric Depression Scale: a executive functions and memory. J Int Neuropsychol comparison with the 30-item form. J Geriatr Psychiatry Soc, 11 (Suppl. 1), 9–10. Neurol 4 (3): 173–8. 56. Amato L, Paolisso G, Cacciatore F, Ferrara N, 62. Sinclair AJ and Woodhouse K (1994) Meeting the Canonico S, Rengo F and Varricchio M. (1996) challenge of diabetes in the aged. J Roy Soc Med 87 Non-insulin-dependent diabetes mellitus is associated (10): 607. with a greater prevalence of depression in the elderly. 28 Depression and Diabetes in Older Adults

Arie Nouwen and Jan R. Oyebode School of Psychology, University of Birmingham, Birmingham, UK

with such a presentation may receive a diagnosis of Key Messages depression. There is currently no objective test to confirm the • Depression in diabetes is a serious comorbidity existence of a depressive disorder, and any diagno- associated with poor outcome and high health sis is based on clusters of signs and symptoms. The care expenditure. International Classification of Diseases (ICD-10) [1] • Clinicians should be alert to signs of depression and the Diagnostic and Statistical Manual of Mental in all older people with diabetes, but especially Disorders-IV (DSM-IV) [2] detail several types of de- those with known risk factors. pression based on symptom duration, the number of • Depression in diabetes can be treated successfully with pharmacotherapy, and/or psycholog- symptoms and the degree of interference with normal ical therapy, but blood glucose levels should activities. The essential features of (uni-polar) depres- be monitored closely, especially with pharma- sion are a persistent lowering of mood, a loss of ability cotherapy. to enjoy usual activities and reduced energy levels (see Table 28.1). While to some degree arbitrary, distinctions 28.1 Defining depression are made between Depressive Episodes/Major Depressive Disorder and Persistent Affective Disor- Feeling ‘depressed’ or ‘low’ does not necessarily indi- ders/Dysthymia. The former is characterized by a loss cate the existence of a depressive disorder. Depressive of pleasure in almost any activity for at least a 2-week symptoms and feelings of unhappiness often occur as period, while dysthymia is a milder but more chronic a transient mood state experienced by almost all inform of depression, with a depressed mood occurring dividuals at some time in their life. These symptoms for most of the day on more days than not, for at and feelings may be due to disappointments, or to dif- least two years. Two other diagnostic categories ficulty in adapting to life events such as bereavement included in ICD-10 that may have relevance for or disability. However, when depressive symptoms ex- people with diabetes are Mild or Moderate Depres- tend across affect, cognition, psychomotor activity and sive Episodes (with mild being equivalent to Minor neurovegetative domains (i.e. appetite, ability to sleep Depressive Disorder in DSM-IV), which involve symp- and sex drive), and interfere with normal function- toms identical to severe depressive episodes but with ing, they are considered pathological and an individual fewer symptoms and less impairment, and Adjustment

Table 28.1 Criteria for a Major Depressive Episode* according to the DSM-IV-R and ICD-10. DSM-IV-R ICD-10 Clinical signiﬁcance Symptoms represent a change from previous functioning Symptoms cause distress or impairment of functioning and cause clinically signiﬁcant distress or impairment in social, occupational, or other important areas of functioning

Duration of symptoms Most of the day, nearly every day for at least 2 weeks Most of the day, almost every day, for at least 2 weeks, largely unaffected by circumstances Severity Five or more symptoms of which at least one a ‘key’ Mild: Four symptoms of which at least two are key symptom symptoms; probably able to continue most activities Mild: ﬁve or six symptoms with mild disability or capacity with increased effort to function normally albeit with substantial effort Moderate: Six symptoms of which at least two are key Severe: presence of most symptoms and clear-cut and symptoms; great difﬁculty in continuing with ordinary observable disability activities Moderate: in between mild and severe Severe: eight symptoms including all three key symptoms; marked distress and impairment Key symptoms (a) Depressed mood most of the day (a) Depressed mood most of the day (b) Diminished interest or pleasure in all or almost all (b) Loss of interest or pleasure in activities that are nor- activities mally pleasurable (c) Decreased energy or increased fatiguability Associated symptoms

1. Significant weight loss or gain or decrease or increase 1. Change in appetite (decrease or increase) with corre- in appetite unrelated to dieting sponding weight change 2. Insomnia or hypersomnia 2. Sleep disturbance 3. Psychomotor agitation or retardation nearly every day 3. Change in psychomotor activity, with agitation or retar- (observable by others) dation 4. Fatigue or loss of energy 4. Unreasonable feelings of self-reproach and inappropriate 5. Feelings of worthlessness or excessive or inappropriate guilt guilt 5. Diminished ability to think or concentrate, such as in- 6. Diminished ability to think or concentrate, or indecisive- decisiveness or vacillation ness 6. Recurrent thoughts of death or suicide, or any suicidal 7. Recurrent thoughts of death, recurrent suicidal ideation behaviour without a specific plan, or a suicide attempt or a specific 7. Loss of confidence or self-esteem plan for committing suicide

∗Note: For a DSM-IV-R diagnosis of Minor Depressive Disorder, a total of at least two but less than ﬁve symptoms is required.

Disorders (ICD-10), which consist of states of Adjustment Disorders in the WHO framework are subjective distress and emotional upset, which similar to the DSM-IV diagnosis of Mood Disorder interfere with social functioning, and arise in the Due to a General Medical Condition, which consists period of adaptation to a significant life change or of a prominent and persistent mood disturbance that as a consequence of a stressful life event (including is judged to be a direct effect of a general medical the presence or possibility of serious physical illness). condition through a physiological mechanism. A key 28.3 EPIDEMIOLOGY OF DEPRESSION IN DIABETES 387 feature of the latter is the presence of a temporal 1.8–2.2). When diagnostic criteria are used (DSM-IV, relationship between the onset, exacerbation, or ICD-10), the pooled point prevalence of depressive remission of the medical condition and the mood disorders has been estimated at about 10%. The pooled disturbance. prevalence of depressive symptoms when self-report One condition that may cause mood symptoms is questionnaires are used is higher at 26%, which cerebrovascular disease (e.g. stroke), and this may have suggests that there are high levels of subthreshold, particular relevance for older adults with diabetes as clinically significant psychological distress in diabetes they are susceptible to cerebrovascular disease [3–6]. populations. In a more recent systematic review and meta-analysis [22], a slightly lower odds ratio was found (OR = 1.59, 95% CI: 1.5–1.7). However, the 28.2 Depression in older people studies included in these meta-analyses were based on cross-sectional data that precluded conclusions about Although growing old is often accompanied by many cause and effect. The finding of an increase in the new and often difficult challenges, depression is not a point prevalence of depression in diabetes has been part of normal ageing [7, 8]. It is estimated that 1–3% reported not only from Western countries but also from of people aged >65 years suffer from major depression various non-Western regions such as Pakistan [23], [9, 10]. These rates are relatively low compared with Iran [24] and Hong-Kong [25]. Although, the level of younger age groups [11], where approximately 15% depression may vary from country to country, people of the general population is at risk of experiencing a with diabetes consistently have higher levels than (major) depressive episode [1, 2]. the general population. However, the prevalence of However, prevalence rates of severe depressive depression in diabetes does not appear to be increased episodes/major depressive disorder (MDD) are higher in comparison with other medical conditions such amongst certain groups of older people, in particular, as heart disease and chronic obstructive pulmonary individuals with a comorbid medical illness, and those disease (COPD) [26, 27]. in residential [12] or nursing homes [13] and in people In addition to an increased prevalence, persistent or with a history of psychiatric illness [14] or previous recurrent clinical and subclinical depression is present depression [15]. in a substantial number of adults with diabetes. In Of relevance to working with diabetes, increased a sample of 245 patients with diabetes [28], depres- depression has also been found in older people with vi- sive symptoms were assessed at the beginning and sual loss (36%) [16] and with cerebrovascular accident end of a comprehensive outpatient diabetes education (CVA) or stroke [17]. Although these numbers con- programme, and at a 6-month follow-up. The rate of cern individuals with a diagnosis of major depression, depression at follow-up was 10% for those without de- much larger numbers report experiencing clinically sig- pressive symptoms at either of the earlier time points. nificant depressive symptoms that do not reach the However, 36% of the people who were depressed at threshold for a clinical diagnosis of severe depressive one earlier time point, and 73% of those depressed at episode or MDD; typically reported patient proportions both earlier time points, reported depressive symptoms are 8–16% [10, 18], 16% [19] and 27% [20]. at follow-up. Likewise, Lustman et al. [29] found that the likelihood of symptomatic affective disorder was only 10% over a 5-year time period in a group of 28.3 Epidemiology of depression in people with diabetes but without depression at time diabetes 1, whereas a recurrence or persistence of clinical depression occurred in 92% of cases with an average of 28.3.1 General population 4.8 depressive episodes over a 5-year period. It is now commonly accepted that the prevalence of depression – that is, the total number of cases in a 28.3.2 Older population population at a given time – is increased in individuals A number of studies have specifically examined de- with diabetes. In a systematic review/meta-analysis of pression in elderly people with diabetes. For example, 42 eligible studies [21], it was found that overall the in one study [30] it was found that while depressive odds of depression in patients with diabetes were dou- symptoms were not more prevalent in an elderly ble those of non-diabetic controls (OR = 2.0, 95% CI: diabetic population, those who had diabetes-related 388 CH 28 DEPRESSION AND DIABETES IN OLDER ADULTS complications were more depressed. Similar results One possible explanation for these divergent results were found in a study conducted in Hong-Kong [25]. is that, in the above-mentioned studies, no differen- This seems to suggest that, when the burden of living tiation was made between early-onset (first episode with diabetes is increased, so too is the level of depres- <60 years) or late-onset (first episode >60 years) sion. However, as these studies were cross-sectional in depression. A large body of research suggests that, design it is impossible to determine whether while the phenomenology of early-onset and late-onset depression increases the chance of developing depression may be indistinguishable [35], late-onset diabetes-related complications, or whether diabetes- depression is more likely to be associated with struc- related complications lead to an increase in depression. tural brain abnormalities [36] and psychosis [37], and More recently, a number of longitudinal studies have less likely to be associated with a family history of examined the incidence (i.e. number of new cases in depression [38]. This suggests different underlying ae- a population) of depression in type 2 diabetes, but tiological processes in depression. the findings have been equivocal. Some studies identified an increased incidence of depression in older 28.4 Depression in diabetes: people with type 2 diabetes. For example, in a prospec- Associated problems tive longitudinal study in a well-functioning cohort of 70–79-year-old persons in the US [31], diabetes 28.4.1 Self-management was found to be associated with a 30% increased risk of incident depressed mood (OR = 1.31; 95% CI: Clinical depression has been associated with poor dia- 1.07–1.61). An adjustment for diabetes-related comor- betes self-care. Both clinical and subclinical depression bidities reduced this relationship (OR = 1.20, 95% CI: were found to be associated with impaired adherence 0.97–1.48). The increased risk for developing depres- to diabetes medication [39, 40] and to anti-depressant sive symptoms was mainly found in those with uncon- medication [41]. This negative association was found even at mild levels of depression [42]. Adjustment trolled diabetes. Participants with a serum HbA1c level >8.1% had the highest risk of developing recurrent for alcohol use, cognitive impairment, age and other depressive symptoms. Using a diagnosis of depression medication use did not significantly alter this nega- rather than self-reported depressive symptoms [32], a tive association. Others have shown that, in elderly large community-based study in Spain found that el- samples, depression is associated with less-frequent derly people with diabetes (age >55 years) had both self-monitoring of blood glucose [42] and dietary a higher prevalence (OR = 1.47; 95% CI: 1.16–1.83) self-care activities [40, 42]. Interestingly, in the latter study it was found that while the consumption of and incidence (OR = 1.40; 95% CI: 1.03–1.90) of de- fruit and vegetables and spacing of carbohydrates were pressive disorder than a non-diabetic control group. affected, the consumption of high-fat foods was not Controlling for potential confounders (demographics, increased. The results of a recent study conducted in cardiovascular risk factors, comorbid medical illness, Korea showed that depressed people with diabetes are cognitive functioning and disability) somewhat atten- also less likely to participate in diabetes education pro- uated these results (prevalent depression: OR = 1.41, grammes [43]. 95% CI: 1.08–1.83; incident depression: OR = 1.26, 95% CI: 0.90–1.77). In contrast, no increase was found in the incidence 28.4.2 Metabolic control of depressive symptoms in a Dutch cohort of people Both clinical and subclinical depression in diabetes aged 55–85 years [33]. Although studying a slightly have been associated with increased symptom younger cohort (age range 50–89 years) in California, reporting [44] and poorer metabolic control. In a Palinkas et al. [34] also reported that diabetes did not meta-analytic review of 24 studies [45], it was found result in a higher incidence of depressive symptoms. that depression was significantly associated with None of the above studies excluded people with hyperglycaemia in both type 1 and type 2 diabetes. a history of depressive disorder. However, as people Larger effect sizes were found when diagnostic criteria with a previous depressive episode are more likely to rather than self-report questionnaires were used, experience further episodes independently of diabetes suggesting that the relationship between depression [29], the results may have been different had the studies and hyperglycaemia is stronger in patients with clinical controlled for previous episodes. rather than subclinical depression. However, the direc- 28.5 POSSIBLE MECHANISMS FOR THE AETIOLOGY OF DEPRESSION 389 tion of this association remained unclear as the studies diabetes self-management, diabetes control and also to included were cross-sectional in nature and precluded reduce complications and health care costs, and even the drawing of any cause-and-effect conclusions. mortality rates.

28.4.3 Diabetes complications 28.5 Possible mechanisms for the Depression has also been associated with increased risk aetiology of depression of developing diabetes-related complications, especially cardiovascular complications [46, 47]. A signif- 28.5.1 Multiple pathways icant association between depression and complication of diabetes, including diabetic retinopathy, nephropa- To date, the aetiology of clinical depression remains thy, neuropathy, macrovascular complications and sex- poorly understood. However, several models have been ual dysfunction, was reported in a meta-analysis of 27 proposed, including genetic, biochemical, psychologi- studies [48]. Effect sizes ranged from small (r = 0.17 cal and social processes and environmental stressors for retinopathy) to moderate (r = 0.32 for sexual dys- [56]. There is consensus that there is no single ex- function). More recently, in a population-based study planation for the pathogenesis of depressive disorder. of older people in Amsterdam [30], the rate of de- Rather, it is believed that the biochemical and psy- pressive disorder was found to be higher in diabetes chosocial changes associated with type 2 diabetes inpatients with related complications than in those with- teract with those of clinical depression [57], resulting out complications. in more severe depression with more frequent and longer-lasting episodes [29, 58]. Yet, the results from 28.4.4 Health care expenditure two large population-based studies – one in the US [59] and one in the Netherlands [60] – have suggested It is not surprising that depression in people with dia- that disturbed glucose homeostasis is not associated betes increases health care expenditure. For example, with depressive symptoms. In fact, in both studies it in a US study it was found that among individuals with was found that people with diabetes, but who were diabetes, the total health care expenditure for individu- unaware of having the condition, did not have an in- als diagnosed with depression was 4.5-fold higher than creased risk of depressive symptoms. Those people di- that for individuals without depression [49]. agnosed with type 2 diabetes (and who therefore were aware of having the condition) had an increased risk 28.4.5 Mortality of developing depressive symptoms compared to subjects with normal glucose concentrations. Also, as the A number of cohort studies have reported increased adjustment for the number of chronic diseases attenu- mortality in people with diabetes [50, 51], including ated the risk of depressive symptoms in patients with in older adults [52]. It is not surprising that the mor- diagnosed diabetes, these results may indicate that de- tality rates are also increased in depressed people. For pressive symptoms in diabetes might be a consequence example, in a longitudinal cohort study [53] it was of its burden rather than a consequence of high glucose found that, over a 2-year period, depression contributed levels. as much to mortality as did myocardial infarction (MI) or diabetes in older primary care patients. The treat- 28.5.2 Stress-vulnerability model ment of depression reduces such risk [54]. By following a large group of people (n = 4154) with diabetes One aetiological model of depression that may account for a period of up to 3 years [55], it was found that for the epidemiological findings described above is a both minor and major depression were strongly asso- stress-vulnerability model. This model posits that in- ciated with increased mortality, even when controlling dividuals having specific depressogenic vulnerabilities for significant predictors of mortality. or defects may suffer depression when these vulner- Although many of the studies cited above were abilities or defects become activated, usually during cross-sectional in nature, thus precluding conclusions stressful situations [61]. When applied to a diabetes about cause and effect, it would seem plausible that the population, this model predicts that the stresses asso- control of depressive symptoms might help to improve ciated with living with diabetes and its ramifications 390 CH 28 DEPRESSION AND DIABETES IN OLDER ADULTS would increase the chances of developing depressive it is natural for an older person who has chronic episodes in people with specific vulnerabilities. health problems to feel somewhat depressed; consequently, the condition is not fully assessed, labelled 28.5.3 Stresses associated with diabetes and treated. The overlap of symptoms of depression with age-related changes (e.g. feeling slowed down; in- Stresses associated with diabetes may vary from sig- terrupted sleep), diabetes-related symptoms (e.g. tired- nificant losses (e.g. bodily functions, amputations, loss ness, loss of libido) and the side effects of medication of vision) to minor hassles (e.g. having to inject in- makes the assessment of depression in older people sulin in social gatherings). At the same time, many very complex. older adults are likely to experience a range of other The aim of an assessment, therefore, is to clearly stresses associated with ageing, including other physical illnesses, loss of roles (e.g. through retirement), establish whether an older person with diabetes is bereavements and social difficulties (e.g. housing or also experiencing depressive symptoms, thoughts or financial problems). All of these have been found to feelings. If so, it becomes important to further assess be associated with greater levels of depression [62]. its extent, severity and impact, so that an appropriate In one study [30] it was found that the rate of de- level of intervention can be provided. pressive disorder was higher in patients with diabetes 28.6.2 The process of assessment complications than in those without complications; this may suggest that a higher burden of diabetes-related A stepped approach is likely to be the most efficient conditions is associated with more depression. It is and effective approach to the identification of depres- also possible that enduring the negative sequelae of sion in diabetes. The steps included comprise a brief stresses related to diabetes accounts for the effects of screening during routine appointments, a full assess- life events on depression [63]. For example, 25% of ment in primary care, and a specialist assessment by patients with diabetes forego social activities because mental health professionals. of social embarrassment related to insulin injections [64]. It is important to note that major life events oc- Brief screening cur quite infrequently, while relatively minor hassles of Of three possible screening instruments used to identify diabetes are probably daily events. Yet daily hassles, depression, Henkel et al. [69] found that the WHO-5 though small in themselves, may lead to a cumulative Well-Being Index had the best sensitivity and speci- impact of stress over time. ficity among 431 adults aged 18–88 years attending Primary Care Services in Germany. Various four-item versions of the Geriatric Depression Scale (GDS) [70, 28.6 The assessment of depression 71] have also been found to be reasonably accurate in older adults with diabetes in the detection of depression specifically among older people with acute physical illness [72, 73] and those in 28.6.1 The aim of assessment primary care [74] (see Table 28.2 for details of these in- Based on the assumption that diabetes will lead to an struments). One of these short scales can easily be built increase in depressed mood, either as a result of the into the conduct of routine appointments with primary biochemical changes directly due to type 2 diabetes, or care staff or specialist diabetes nurses. resulting from the burden of living with diabetes [57], Assessment of depression in primary care it is not surprising that clinical guidelines (e.g. [65–67]) now recommend that all patients with diabetes undergo Where a simple brief screen indicates possible depres- regular screening for depression. In particular, atten- sion, then a fuller assessment should be carried out tion must be paid to the routine screening of those which may lead to either intervention within primary with known vulnerability factors, such as a previous care or referral on to mental health services for spe- history of depression, diabetes-related complications, cialist assessment and treatment. poorly controlled diabetes, diagnosis of dementia and The assessment should preferably combine clini- psychological stress. cal observation, information about the patient’s his- There is evidence that depression among older peo- tory, current functioning, thoughts and feelings and ple attending primary care health services goes un- a standardized questionnaire. High levels of success- detected [68], perhaps because it is assumed that ful suicide among older people with multiple physical 28.6 THE ASSESSMENT OF DEPRESSION IN OLDER ADULTS WITH DIABETES 391

Table 28.2 Screening Tools for Depression. The WHO-5 Well-Being Index Respondents are asked to indicate the frequency of their experience in relation to each of five questions over the past 2 weeks on a six-point Likert scale which ranges from ‘All of the time’ to ‘At no time’. 1. I have felt cheerful and in good spirits 2. I have felt calm and relaxed 3. I have felt active and vigorous 4. I woke up feeling fresh and rested 5. My daily life has been filled with things that interest me 4-item versions of the Geriatric Depression Scale Respondents are asked to answer ‘yes’ or ‘no’ to whether they have felt like this over the past week. 1. Are you basically satisfied with your life? 2. Do you feel that your life is empty? 3. Are you afraid that something bad is going to happen to you? 4. Do you feel happy most of the time? [72, 131, 132] Or 1. Are you basically satisfied with your life? 2. Have you dropped many of your activities and interests? 3. Do you feel happy most of the time? 4. Do you prefer to stay at home, rather than going out and doing new things? [74] illnesses [75] mean that questions about the presence Shorter versions have also been developed with of suicidal ideas should always be asked if moderate 15 items [76, 77], ten items, four items and one to severe depression is present. item [73]. The 15-item version has been found to be highly correlated with the total score from the longer 28.6.3 Measures specific to older people: version, and to have high internal consistency, good The geriatric depression scale sensitivity and specificity among 198 older people One of the most widely used measures for depres- attending primary care services [73]. Both sensitivity sion in older people is the Geriatric Depression Scale and specificity fall when even shorter versions are used (GDS) [70, 71]. The original version has 30 items con- [72, 73] although, as noted above, briefer versions may cerning thoughts and feelings typical of depression in still be helpful in initial screening. older people, and the respondent is asked to answer Scores on the GDS correlate highly with clinical yes or no to whether they have experienced these dur- diagnoses of depression [78]. Various research studies ing the past week. The questions do not include the have indicated that the GDS is applicable across a somatic and physical symptoms that may fail to dis- wide variety of clinical populations, including primary criminate depression from other conditions in older care settings [74], medical outpatients and inpatients, people. The yes/no scale is easier for many older peo- nursing home residents and older people in mental ple to answer than the four- or five-point Likert scales health services [79]. A recent review of assessment that are used in general adult questionnaires, and there for depression in older people identified nine studies is also an informant version. The scheme is not sub- evaluating use of the 15- and 30-item GDS systems in ject to copyright, and is readily available from the primary care settings [80]. These reported that sensitiv- Internet. ity and specificity ranged from 79–100% and 67–80%, 392 CH 28 DEPRESSION AND DIABETES IN OLDER ADULTS respectively. The GDS has also been specifically symptoms in diabetes research [52]. The CES-D is evaluated and found to be valid in a British often used in older populations [91–93], and has been African-Caribbean patient sample [81, 82]. shown to have wide cross-cultural use in the USA with Black and White older adults [91]. 28.6.4 Measures for people with dementia The CES-D consists of 20 items answered on a 4-point Likert scale, and enquires about symptoms ex- Whilst the GDS-15 may be the scale of choice for perienced over the past week. A review of its prop- older people without cognitive impairment, it does not erties in use with older populations found it to have perform as well in those with dementia. A direct com- sensitivity and specificity ranging from 75–93% and parison of the GDS with the Cornell Scale for Depres- 73–87%, respectively, for the detection of severe de- sion in Dementia (CSDD) [83] in a sample of patients pressive episodes/major depression. However, it may aged >65 years, which included people with depres- be less effective for identifying minor, subclinical lev- sion alone, dementia alone, depression and dementia els of depression [94]. These authors noted that the and neither, found the CSDD to retain its specificity CES-D was reported as being difficult to administer to and sensitivity among people with depression and de- those in nursing homes, where the respondents found mentia better than the GDS [84]. it difficult to remember and answer the questions. Al- The CSDD involves collecting information from though the sensitivity of the CES-D to detect clinical both the older person with dementia and an informant. depression was lower than that of the BDI, its speci- It includes 19 items on mood, physical and behavioural ficity was higher [90]. indicators, ideological symptoms and changes in diurnal patterns, with a three-point response scale. It has 28.6.6 Summary been used in samples of older people in both nursing homes [85] and community settings [84], and should Clinicians should be alert to the signs of depression be used in place of the GDS when it is thought that the in all older people with diabetes, but especially those subject might have cognitive impairment (e.g. a Mini with known risk factors. For these people, brief Mental State Examination score <27). screening questions – as listed in Table 28.2 – should be routinely included in primary care appointments. Where these indicate possible depression, or where 28.6.5 General population measures this is otherwise suspected, a fuller assessment The most common general population measures of should be carried out in primary care or, where depression in diabetes research and practice include there are complicating factors, in specialist mental the Centre for Epidemiological Studies Depression health services. For those patients without significant scale (CES-D) [86] and the Beck Depression Inventory cognitive impairment, the 15-item version of the (BDI) [87]. GDS seems to be the instrument of choice to assess The BDI is a very widely used and readily available depression in primary care, hospital or residential scale, and the indications are that it is probably valid settings. The CSDD should be used in preference with older populations. However, it is a little longer where the diabetic patient has significant cognitive (21 items) and is more complex (4-point Likert scale) deficits. However, where data are being collected, or to complete than the GDS. Research with the BDI has need to be compared across the adult age range, then shown that items concerned with physiological signs of the BDI or CES-D may be preferable. Where scores depression are minimally confounded with symptoms indicate possible depression, and this is supported by of diabetes [88], and it was found that eliminating the clinical information and observation, treatment can somatic symptoms did not improve the accuracy of this be undertaken in primary care, or the patient may measure [89]. Nonetheless, the BDI was found to have be referred to a specialist mental health service. The satisfactory sensitivity and specificity to detect clinical guidelines issued by organizations such as the UK depression in a mixed sample of type 1 and type 2 National Institute for Health and Clinical Excellence diabetes patients [90]. [95] suggest that mild depression be treated in primary The CES-D has been used in a wide range of people care, whilst moderate to severe depression may be in different situations and settings, and is now the treated in either primary or secondary care, depending most commonly used self-report measure of depressive on the patient’s particular circumstances and history. 28.7 THE TREATMENT OF CLINICAL DEPRESSION 393 28.7 The treatment of clinical optimal therapeutic effect. A more recent Cochrane review confirmed that SSRIs and TCAs were of sim- depression ilar efficacy, but that TCAs had a higher withdrawal rate due to their side effects [99]. The goals for treating depressed diabetes patients are In all age groups there is evidence that patients with two-fold: (i) remission or improvement of depressive depression and comorbid medical illness are signifi- symptoms; and (ii) improvement of the often poor cantly more likely to improve if treated with antide- glycaemic control, which is generally considered as pressants compared to placebo or no treatment [100, fundamental to the management of diabetes, in order 101]. Specifically, however, although the TCA nor- to prevent or delay long-term complications. triptyline was shown to improve depression, it was The treatment of depression may have a positive more likely to worsen the control of diabetes [100, effect on levels of motivation to manage activities 101] and to have an adverse effect on glucose con- of daily living, to take physical exercise, and to ad- trol [102]. In fact, TCAs have been shown to increase here to dietary management and medication regimes. serum glucose levels by up to 150%, causing hyper- However, only a handful of randomized control trials glycaemia as well as increasing appetite and reducing have been conducted examining the treatment of de- metabolic rate, resulting in weight gain [102–105]. In pression in people with diabetes, and these have been view of the above, TCAs such as nortriptyline would small-scale in design. Even fewer studies have specif- not be the first-line treatment for depression in diabetes. ically targeted an older population. However, the re- Consistent with the guidance for treatment of older sults to date appear to suggest that cognitive behaviour adults cited above, in those with diabetes SSRIs have therapy and antidepressant medication are as effec- been shown to be effective, with most data supporting tive in people with major depressive disorder (MDD) the use of fluoxetine, as this improves both mood and and diabetes as in those with MDD without diabetes glycaemic control [106, 107]. There is also promis- with, in some instances, additional beneficial effects ing evidence of the efficacy of sertraline [108, 109]. on glycaemic control [96]. Older depressed primary Fluoxetine improves HbA levels, reduces insulin re- care patients with diabetes in practices that implement 1c quirements, and is associated with weight loss and depression care management [monitoring, antidepres- enhanced insulin sensitivity [110]. Moreover, its ef- sants such as serotonin reuptake inhibitors (SSRIs) or fects on insulin sensitivity are independent of its ef- inter-personal therapy and follow-up sessions] have fect on weight loss [111]. Despite these advantages, it been found to have lower mortality levels over the has been reported that patients with diabetes may be- course of a 5-year interval than depressed patients with come hypoglycaemic during treatment with fluoxetine diabetes in usual-care practices (adjusted hazard ratio [112]. Furthermore, the side effects of fluoxetine in- 0.49, 95% CI: 0.24–0.98) [54]. clude tremor, nausea, sweating and anxiety, and these may be mistaken by patients for an episode of hypo- 28.7.1 Pharmacotherapy glycaemia. It has therefore been advised that patients According to the expert consensus guideline on the are informed of these possible effects [113]. treatment of depressive disorders in older people [97], SNRIs such as venlafaxine do not appear to disrupt it is important to identify coexisting medical condi- glycaemic control, although only limited data are avail- tions, such as diabetes mellitus, which may contribute able on the efficacy of this drug [103]. Duloxetine has to depression itself or complicate its treatment. The been shown to be effective in the treatment of dia- preferred first-line treatment is a SSRI or a serotonin betic peripheral neuropathic pain and, given that it is norepinephrine reuptake inhibitor (SNRI), and psy- also an antidepressant, it would be a suitable treatment chotherapy [97]. A Cochrane Database systematic re- for depression in diabetes. It has little influence on gly- view on the use of antidepressant versus placebo treat- caemic control, but has been shown to be both safe and ment in depressed elderly people [98] found that SS- well-tolerated, with few adverse side effects [114]. RIs, tricyclic antidepressants (TCAs) and monoamine Other available antidepressants include MAOIs reuptake inhibitors (MAOIs) were all effective for the and newer agents (mirtazapine, reboxetine and treatment of depression in older people, and empha- trazadone). Unfortunately, the MAOIs have been sized the need to wait for at least 6 weeks to ensure found to cause not only weight gain [108, 115] but 394 CH 28 DEPRESSION AND DIABETES IN OLDER ADULTS also hypoglycaemic episodes [108], and there is work is provided in primary care. Where further in- little evidence as yet of the effectiveness of these tervention is indicated, higher intensity psychological newer agents [116, 117]. Were either to be used, they therapy is then recommended either in primary or would probably best be administered under specialist secondary care [123]. There is some evidence for the psychiatric care. effectiveness of both these levels of psychological in- In all patients with diabetes being treated with an- tervention for depression in diabetes. tidepressants, it is important to monitor blood sugar In primary care, interventions such as ‘depression levels closely at the start of treatment, when the dose care management’ (which consists of patient education, in changed, or after discontinuation [104]. support with adherence to medication, and assistance with coordination of primary and secondary care) [124] 28.7.2 Maintenance therapy and ‘depression case management’ (education, medication support or problem-solving therapy) [125] have Current clinical guidelines [118] recommend that, fol- been found to reduce depression and increase exercise lowing a first episode of depression, treatment should [124], or to reduce depression and increase satisfac- continue for 4–6 months. Maintenance treatment be- tion with care [125], with effects continuing at the yond 6 months is recommended for individuals who 12-month follow-up. have had two or more episodes, have a comorbid phys- Higher intensity psychological interventions such as ical illness, or have failed to make a complete recovery counselling [126] and cognitive behavioural therapy from their depressive episode. Other risk factors to take (CBT) [121] have provided positive results regarding into consideration include age and ongoing psychoso- the improvement of depression, and have shown mod- cial difficulties. erate to good positive effects with regards to glycaemic A number of studies have focused specifically on control. those with diabetes. In patients with diabetes, mainte- A recent systematic review and meta-analysis of the nance therapy with sertraline has been found to prolong effectiveness of psychological interventions (including the depression-free interval following recovery from counselling, CBT and psychodynamic psychotherapy) major depression [109]. Depression recovery with ser- to improve glycaemic control in those with type 2 traline, as well as sustained remission with or without diabetes [127], concluded that these therapies led to treatment, is associated with improvements in HbA1c long-term improvements in psychological distress and levels for at least 1 year [109]. in glycaemic control. However, it should be noted that A recent trial of maintenance pharmacotherapy for the studies included were not aimed primarily at the the prevention of depression recurrence showed a ben- treatment of depression. efit of sertraline in younger patients with diabetes (aged <55 years) compared to placebo in patients who ini- 28.7.4 Treatment for mild to moderate tially responded to sertraline, but not in the older pop- depression ulation (>55 years). However, this was mainly due to a higher response to non-specific treatment factors in Studies exploring the effects of treatment of mild this group [119]. to moderate depression in diabetic patients are rare, Although pharmacotherapy is the most likely pre- although this subthreshold form of depression is highly scribed treatment for MDD in primary care [120], it prevalent in the diabetic population [21] and increases should be noted that in 50% of depressed diabetes the risk of subsequent major depression [128]. To date, patients pharmacotherapy does not lead to a full re- only a small randomized, placebo-controlled trial of mission of depression [58, 121, 122]. Therefore, new paroxetine in mildly depressed postmenopausal women approaches to the treatment of depression in diabetes with type 2 diabetes (age >50 years) [129] has been seem warranted. carried out, and has shown beneficial treatment effects on insulin sensitivity and glycaemic control. Self-help strategies for early intervention against 28.7.3 Psychological interventions subclinical depression, though proven to be beneficial Psychological interventions may be provided within a in the general population [130], remain to be examined stepped care model in which self-help or low-intensity in persons with diabetes. 28.8 CASE STUDIES 395 28.8 Case studies (depressed mood, loss of interest and enjoyment and reduced energy) as well as a number of others (re- 28.8.1 Case study 1 duced self-esteem and self-confidence, ideas of guilt, a bleak and pessimistic view of the future and disturbed Presentation sleep). This fulfils the criteria for a diagnosis of severe EH, a woman in her mid-70s who lives alone, is well depressive episode. known to the Practice having a history of both physical and mental health problems. Her postnatal depression Treatment many years ago was treated by her GP and a later In view of her score on the GDS-15, her clinical pic- episode of depression, after her husband died 20 years ture and her history, the GP offers EH antidepressants ago, was treated through inpatient psychiatric care. and also refers her to the specialist Mental Health Ser- She has had no problems with depression since that vice for further assessment and treatment. That Service time. However, she has some chronic health problems supports the introduction of the antidepressants and is including heart failure treated with diuretics, COPD able to arrange for her to have a series of appointments and type 2 diabetes treated through medication. Her for cognitive-behavioural therapy (CBT). The CBT diabetes is poorly controlled. She attends the surgery therapist focuses with her on: (i) combating learned looking thin, frail and unkempt. helplessness through re-introducing pleasant activities; (ii) helping her to learn to challenge negative think- Commentary The indicators of possible depression ing about her lack of control over her health and her on this occasion are EH’s lack of self care and loss life; (iii) helping her to build self-esteem by contact- of weight, although her poorly controlled diabetes ing friends; and (iv) helping her to consider alternative may also contribute to the weight loss. Taking these ways that her friends might see her dietary needs. The in conjunction with vulnerability factors (her known GP also refers her to a local programme of Physical history of depression and her physical health problems) Rehabilitation for COPD as exercising reduces depres- an assessment for possible depression is warranted. sive symptoms as well as improving diabetes control and COPD. As a result of these interventions, EH’s Assessment depression improves, her sense of guilt reduces and Her GP sees EH at the surgery and asks how she is. her optimism returns, along with better self-care and She says that over the past 2–3 weeks, she has been dietary management. sleeping badly and is too breathless to go out, so is confined to the house. She cannot bring herself to put Commentary In line with clinical guidelines, EH’s effort into cooking and cleaning as it seems pointless treatment involves both medication and psychologi- when she cannot do it to her usual high standard. cal therapy. Typical of work with older people, a She complains that the days seem very long being at multi-pronged approach is used to address a range of home on her own. She no longer finds any pleasure in physical health (lack of exercise), mental health (nega- watching TV or reading. She does not ring friends as tive thinking) and social (lack of social contact) issues. she feels she is not good company and her diet will Following her initial improvement, consideration will be a trouble to them. She cannot see things getting now need to be given to its maintenance. better and expresses the view that her current state is a punishment for an extra-marital affair she had much 28.8.2 Case study 2 earlier in life. She is convinced that no-one can do anything to improve her health and sometimes wonders Presentation if life is worth living. BC is a 65-year-old recently retired plumber who was diagnosed with type 2 diabetes at the age of Commentary EH gives evidence of having seven 52. He manages his diabetes through insulin (two of the symptoms of depression that are outlined in injections per day), but is considerably overweight. He ICD-10, including the three that are viewed as core has hypertension, which is currently being controlled 396 CH 28 DEPRESSION AND DIABETES IN OLDER ADULTS

with medication, and has sustained diabetes-related Commentary In line with best practice, BC’s mild damage to his vision. On a recent visit to the surgery depression is dealt with in Primary Care, through brief he complains of lack of energy, a pervasive sense of therapy focused on helping him to help himself through tiredness and a loss of interest in his usual past-times. establishing satisfying activities and ﬁnding solutions to current problems. His loss of weight subsequent to Commentary BC’s fatigue and withdrawal may be joining the walking group may help in the longer term linked directly to physical health, but could equally be control of his diabetes as well as in keeping depression manifestations of depression. away.

Assessment 28.8.3 Case study 3 The Practice Nurse becomes aware of BC’s situation Presentation and state of mind when she sees him for routine DW is a woman in her mid-80s who has mild vascular monitoring of blood pressure. She books a double dementia and diabetes, managed through dietary con- appointment time with him for the following week trol. She has recently moved into a Residential Home when she establishes more about his current thoughts as she was unable to look after herself at home, but and feelings and completes the GDR-15. staff are finding her difficult to manage. She is not mixing with others and her personal hygiene is poor. She Commentary BC’s score on the GDS-15 is 7, which shouts aggressively at staff when they try to encourage is at the lower end of the range and suggestive of mild her to wash. depression. He is complaining of the three core symptoms of depression (reduced energy, loss of interest and Commentary Withdrawal, lack of self-care and irri- enjoyment and depressed mood) as well as disturbed tability may all be signs of depression or of dementia. sleep and reduced self-esteem and self-confidence. This In addition, Miss W has a number of vulnerability is sufficient for him to receive a formal diagnosis of a factors for depression (diabetes, dementia and in Res- mild depressive episode. idential Care).

Treatment Assessment The Practice Nurse and GP in consultation decide to The Diabetes Specialist Nurse, who has known DW for ask the Primary Care Mental Health Worker (PCMHW) some time, visits the Home to find out more from the who is based in the Practice to see BC. She sees him for staff and to ascertain how Miss W feels she is settling six appointments over the following 10 weeks. She and in and managing. She fills out the CSDD and discovers he agree that his current feelings of uselessness may a total score of 21. be exacerbated by his recent retirement and his deteriorating vision. They use a problem-solving approach Commentary The score of 21 in the CSDD is within and elements of CBT to consider ways of finding pur- the range typical of the presence of definite ‘major pose and structure in life following retirement. As a depression’. result, BC turns the manual dexterity he formerly used in plumbing to use in a new hobby of model build- Treatment ing. The PCMHW also directs him to a local charity that offers expert assessment and loans of aids to those The Specialist Nurse refers Miss W to the Commu- with low vision. He visits the charity, and as a result nity Mental Health team for Older People. The As- joins a walking group designed to build confidence in sociate Specialist in the team prescribes appropriate those with acquired visual impairment. This gives him anti-depressants for DW. The CPN provides the staff exercise, enables him to lose weight and introduces of the Home with information about depression in de- him to new acquaintances, one of whom becomes a mentia and helps them to work with DW to draw up good friend. Following the intervention, BC reports a person-centred care plan that involves reconnecting feeling more energetic and more satisfied with his life. her with some of her friends in the community. These His score on the GDS-15 has reduced to 2, which is a measures lead to a relief of her depression and DW level typical of those who are not depressed. becomes less irritable. 28.9 SUMMARY 397

Commentary As with EH, treatment involves more 11. Weissman MM, Leaf PJ, Tischler GL, Blazer DG, than one approach, in this case medication and psy- Karno M, Bruce ML, et al. Affective disorders in five chosocial intervention. The backdrop of cognitive im- United States communities. Psychol Med. 1988; 18 pairment means that the intervention is carried out with (1): 141–53. the collaboration of those staff who are in day-to-day 12. Ames D. Epidemiological studies of depression contact with the patient. among the elderly in residential and nursing homes. Int J Geriatr Psychiatry 1991; 6, 347–54. 13. Parmalee P, Katz I, Lawton M. The design of spe- 28.9 Summary cial environments for the aged. In: J. Birren and K. Schaie (eds), Handbook of the Psychology of Aging, In summary, there have been encouraging findings 3rd edition. Academic Press, San Diego; 1992, pp. that depression can be treated successfully in diabetes 464–88. patients, with at least good short-term results. As in 14. Van Ojen R, Hooijer C, Jonker C, Lindeboom J, van the general population, there is no evidence that either Tilburg W. Late-life depressive disorder in the com- pharmacotherapy or psychological therapy is superior munity, early onset and the decrease of vulnerability to the other, and the choice of treatment may ultimately with increasing age. J Affect Disord. 1995; 33 (3), 159–66. depend on patient preference. 15. Cole MG, Dendukuri N. Risk factors for depression among elderly community subjects: a systematic review and meta-analysis. Am J Psychiatry 2003; 160: References 1147–56. 16. Zarit S, Zarit J. Mental Disorders in Older Adults: 1. World Health Organisation. ICD-10 Classification of Fundamental of Assessment and Treatment. Guilford, Mental and Behavioural Disorders. World Health Or- New York; 1998. ganisation, Geneva; 1992. 17. Sharpe M, Hawton K, Seagroatt V, Bamford J, House 2. American Psychiatric Association. Diagnostic and A, Molyneux A, et al. Depressive disorders in long Statistical Manual of Mental Disorders, Fourth Edi- term survivors of stroke: associations with demo- tion. Washington, DC: American Psychiatric Associ- graphic and social factors, functional status and brain ation; 1994. lesion volume. Br J Psychiatry 1994; 164 (3), 380–6. 3. Alexopoulos GS, Meyers BS, Young RC, Campbell S, 18. National Institute of Health Consensus Development Silbersweig D, Charlson M. ‘Vascular depression’ hy- conference. Diagnosis and treatment of depression in pothesis. Arch Gen Psychiatry 1997; 54 (10): 915–22. late life. JAMA 1992; 268 (8): 1018–24. 4. Bruce DG, Casey G, Davis WA, Starkstein SE, Clar- 19. Livingstone G, Hawkins A, Blizard B, Mann A. The nette RC, Foster JK, et al. Vascular depression in Gospel Oak study: prevalence rates of dementia, de- older people with diabetes. Diabetologia 2006; 49 pression and activity limitation among elderly resi- (12): 2828–36. dents in inner London. Psychol Med. 1990; 20 (1): 5. Krishnan KR, Hays JC, Blazer DG. MRI-defined 137–46. vascular depression. Am J Psychiatry 1997; 154 (4): 20. Blazer D, Hughes D, George L. The epidemiology 497–501. of depression in an elderly community population. 6. Robinson RG. Poststroke depression: prevalence, di- Gerontologist 1987; 27: 281–7. agnosis, treatment, and disease progression. Biol Psy- 21. Anderson RJ, Freedland KE, Clouse RE, Lustman PJ. chiatry 2003; 54 (3): 376–87. The prevalence of comorbid depression in adults with 7. Lewinsohn PM, Rohde P, Seeley JR, Fischer SA. Age diabetes: a meta-analysis. Diabetes Care 2001; 24 (6), and depression: unique and shared effects. Psychol 1069–78. Aging 1991; 6(2): 247–60. 22. Ali S, Stone MA, Peters JL, Davies MJ, Kunti K. 8. Roberts RE, Kaplan GA, Shema SJ, Strawbridge WJ. The prevalence of co-morbid depression in adults Does growing old increase the risk for depression? with type 2 diabetes: a systematic review and Am J Psychiatry 1997; 154 (10): 1384–90. meta-analysis. Diabet Med. 2006; 23 (11): 1165–73. 9. Chapman DP, Perry GS. Depression as a major com- 23. Zahid N, Asghar S, Claussen B, Hussain A. Depres- ponent of public health for older adults. Prev Chronic sion and diabetes in a rural community in Pakistan. Dis. 2008; 5(1): http://www.cdc.gov/pcd/issues/2008/ Diabet Res Clin Prac. 2008; 79 (1): 124–7. jan/07 0150.htm (accessed December 2007). 24. Khamseh ME, Baradaran HR, Rajabali H. Depression 10. Blazer D. The epidemiology of depression in late life. and diabetes in Iranian patients: a comparative study. J Geriatr Psychiatry 1989; 22: 35–52. Int J Psychiatry Med. 2007; 37 (1): 81–6. 398 CH 28 DEPRESSION AND DIABETES IN OLDER ADULTS

25. Chou KL, Chi I. Functional disability related to di- 37. Kessing LV. Differences in diagnostic subtypes abetes mellitus in older Hong Kong Chinese adults. among patients with late and early onset of a single Gerontologist 2005; 51: 334–9. depressive episode. Int J Geriatr Psychiatry. 2006; 21 26. Egede LE. Major depression in individuals with (12): 1127–31. chronic medical disorders: prevalence, correlates and 38. Mendlewicz J. The age factor in depressive illness: association with health resource utilization, lost pro- some genetic considerations. J Gerontol. 1976; 31 (3): ductivity and functional disability. Gen Hosp Psychi- 300–3. atry 2007; 29 (5): 409–16. 39. Kalsekar ID, Madhaven SS, Amonkar MM, Makela 27. Patten SB, Beck CA, Kassam A, Williams JV, Bar- EH, Scott VG, Douglas SM, et al. Depression in bui C, Metz LM. Long-term medical conditions and patients with type 2 diabetes: impact on adherence to major depression: strength of association for specific oral hypoglycemic agents. Ann Pharmacother. 2006; conditions in the general population. Can J Psychiatry 40 (4): 605–11. 2005; 50 (4): 195–202. 40. Ciechanowski PS, Katon WJ, Russo JE. Impact of 28. Peyrot M, Rubin RR. Persistence of depressive symp- depressive symptoms on adherence, function, and toms in diabetic adults. Diabetes Care 1999; 22 (3): costs. Arch Intern Med. 2000; 160: 3278–85. 448–52. 41. Kilbourne AM, Reynolds CF, Good CB, Sereika SM, 29. Lustman PJ, Griffith LS, Clouse RE. Depression in Justice AC, Fine MJ. How does depression influence adults with diabetes. Results of 5-yr follow-up study. diabetes medication adherence in older patients? Am Diabetes Care 1988; 11 (8): 605–12. J Geriatr Psychiatry. 2005; 13 (3): 202–10. 30. Pouwer F, Beekman AT, Nijpels G, Dekker JM, Snoek 42. Gonzalez J, Safren S, Cagliero E, Wexler D, Dela- FJ, Kostense PJ, et al. Rates and risks for co-morbid hanty L, Wittenberg E, et al. Depression, self-care, depression in patients with Type 2 diabetes mellitus: and medication adherence in type 2 diabetes: relationships across the full range of symptom severity. results from a community-based study. Diabetologia Diabetes Care 2007; 30 (9): 2222–7. 2003; 46 (7): 892–8. 43. Park H, Hong Y, Lee H, Ha E, Sung Y. Individuals 31. Maraldi C, Volpato S, Penninx BW, Yaffe K, Si- with type 2 diabetes and depressive symptoms exhib- monsick EM, Strotmeyer ES, et al. Diabetes mellitus, ited lower adherence with self-care. J Clin Epidemiol. glycemic control, and incident depressive symptoms 2004; 57 (9): 978–84. among 70- to 79-year-old persons: the health, aging, 44. Ciechanowski PS, Katon WJ, Russo JE, Hirsch IB. and body composition study. Arch Intern Med. 2007; The relationship of depressive symptoms to symptom 167 (11): 1137–44. reporting, self-care and glucose control in diabetes. 32. De Jonghe P, Roy J, Saz P, Marcos G, Lobo A, Gen Hosp Psychiatry 2003; 25: 246–52. ZARADEMP investigators. Prevalent and incident 45. Lustman PJ, Anderson RJ, Freedland KE, de Groot depression in community dwelling elderly persons M, Carney RM, Clouse RE. Depression and poor with diabetes mellitus: results from the ZARADEMP glycemic control: a meta-analytic review of the lit- project. Diabetologia 2006; 49 (11): 2627–33. erature. Diabetes Care 2000; 23 (7): 934–42. 33. Bisschop MI, Kriegsman DM, Deeg DJ, Beekman 46. Carney R, Freedland K, Lustman P, Griffith L. De- AT, van Tilburg W. The longitudinal relation between pression and coronary disease in diabetic patients: a chronic diseases and depression in older persons in 10-year follow-up (Abstract). Psychosom Med. 1994; the community: the Longitudinal Aging Study Ams- 56: 149. terdam. J Clin Epidemiol. 2004; 57: 187–94. 47. Lloyd CE, Kuller LH, Ellis D, Becker DJ, Wing RR, 34. Palinkas LA, Lee PP, Barrett-Connor E. A prospective Orchard TJ. Coronary artery disease in IDDM. Gender study of Type 2 diabetes and depressive symptoms in differences in risk factors but not risk. Arterioscler, the elderly: the Rancho Bernardo Study. Diabet Med. Thromb Vasc Biol. 1996; 16 (6): 720–6. 2004; 21 (11): 1185–91. 48. de Groot M, Anderson R, Freedland KE, Clouse 35. Janssen J, Beekman AT, Comijs HC, Deeg DJ, Heeren RE, Lustman PJ. Association of depression and di- TJ. Late-life depression: the differences between abetes complications: a meta-analysis. Psychosom early- and late-onset illness in a community-based Med. 2001; 63 (4): 619–30. sample. Int J Geriatr Psychiatry 2006; 21 (1): 86–93. 49. Egede LE, Zheng D, Simpson K. 2002 Comorbid 36. Simpson S, Baldwin R, Jackson A, Burns A, Thomas depression is associated with increased health care P. Is the clinical expression of late-life depression use and expenditures in individuals with diabetes. influenced by brain changes? MRI subcortical neu- Diabetes Care 2002; 25 (3): 464–70. roanatomical correlates of depressive symptoms. Int 50. Zhang X, Norris SL, Gregg EW, Cheng YJ, Beckles Psychogeriatr. 2000; 12 (4): 425–34. G, Kahn HS. Depressive symptoms and mortality 28.9 SUMMARY 399

among persons with and without diabetes. Am J 65. National Institute of Clinical Excellence. Depression. Epidemiol. 2005; 161 (7): 652–60. Management of depression in Primary and Secondary 51. Egede LE, Nietert PJ, Zheng D. Depression and Care. Clinical Guideline 23. NICE, London, 2004. all-cause and coronary heart disease mortality among 66. American Diabetes Association. Standards of Medical adults with and without diabetes. Diabetes Care 2005; Care in Diabetes. Diabetes Care 2005; 28 (Suppl. 1): 28 (6): 1339–45. S4–36. 52. Black SA, Markides KS, Ray LA. Depression pre- 67. International Diabetes Federation. Global Guideline dicts increased incidence of adverse health outcomes for type 2 diabetes. International Diabetes Federation, in older Mexican Americans with type 2 diabetes. Di- Brussels, Belgium; 2005, pp. 19–21. abetes Care 2003; 26: 2822–8. 68. Iliffe S, Haines A, Gallivan S, Booroff A, Goldenberg 53. Gallo JJ, Bogner HR, Morales KH, Post EP, Ten E, Morgan P. Assessment of elderly people in general Have T, Bruce ML. Depression, cardiovascular dis- practice. 1. Social circumstances and mental state. Br ease, diabetes, and two-year mortality among older, J Gen Pract. 1991; 41 (342), 9–12. primary-care patients. Am J Geriatr Psychiatry. 2005; 69. Henkel V, Mergl R, Kohnen R, Maier W, Moller¨ 13 (9): 748–55. HJ, Hegerl U. Identifying depression in primary care: 54. Bogner HR, Morales KH, Post EP, Bruce ML. Dia- a comparison of different methods in a prospective betes, depression, and death: a randomized controlled cohort study. Br Med J. 2003; 326 (7382): 200–1. trial of a depression treatment program for older adults 70. Brink T, Yesavage J, Lum O, Heersuma P, Adey based in primary care (PROSPECT). Diabetes Care M, Rose T. Screening tests for geriatric depression. 2007; 30: 3005–10. Clinical Gerontologist 1982; 1: 37–43. 55. Katon WJ, Rutter C, Simon G, Lin EH, Ludman 71. Yesavage J, Brink T, Rose T, Lum O, Huang V, Adey E, Ciechanowski P, Kinder L, et al. The association M, et al. Development and validation of a geriatric depression screening scale: a preliminary report. J of comorbid depression with mortality in patients Psychiatric Res. 1983; 17 (1): 37–49. with type 2 diabetes. Diabetes Care 2005; 28 (11), 72. Shah A, Herbert R, Lewis S, Mehendran R, Platt 2668–72. J, Bhattacharyya B. Screening for depression among 56. Southwick SM, Vythilingam M, Charney DS. The acutely ill geriatric inpatients with a short Geriatric psychobiology of depression and resilience to stress: Depression Scale. Age Ageing 1997; 26 (3): 217–21. implications for prevention and treatment. Annu Rev 73. D’Ath P, Katona P, Mullan E, Evans S, Katona C. Clin Psychol. 2005; 1: 255–91. Screening, detection and management of depression 57. Talbot F, Nouwen A. A review of the relationship in elderly primary care attenders. I: The acceptability between depression and diabetes in adults: is there a and performance of the 15 item Geriatric Depression link? Diabetes Care 2000; 23 (10): 1556–62. Scale (GDS15) and the development of short versions. 58. Lustman PJ, Grifﬁth LS, Freedland KE, Clouse RE. Fam Pract. 1994; 11 (3): 260–6. The course of major depression in diabetes. Gen Hosp 74. van Marwijk H, Wallace P, de Bock G, Hermans J, Psychiatry 1997; 19 (2): 138–43. Kaptein A, Mulder JD. Evaluation of the feasibility, 59. Palinkas LA, Barrett-Connor E, Wingard DL. Type reliability and diagnostic value of shortened versions 2 diabetes and depressive symptoms in older adults: of the geriatric depression scale. Br J Gen Pract. 1995; a population-based study. Diabet Med. 1991; 8 (6): 45 (393): 195–9. 532–9. 75. Juurlink D, Herrmann N, Szalai J, Kopp A, Re- 60. Knol MJ, Heerdink ER, Egberts AC, Geerlings MI, delmeier D. Medical illness and the risk of suicide Gorter KJ, Numans ME, et al. Depressive symptoms in the elderly. Arch Intern Med. 2004; 164 (11), in subjects with diagnosed and undiagnosed type 2 1179–84. diabetes. Psychosom Med. 2007; 69 (4): 300–5. 76. Sheikh J, Yesavage J. Geriatric Depression Scale: 61. Banks SM, Kerns RD. Explaining high rates of de- recent ﬁndings and development of a short version. pression in chronic pain: A Diathesis-stress frame- In: T. Brink (ed.). Clinical Gerontology: A guide work. Psychological Bulletin 1996; 119, 95–110. to assessment and intervention. Howarth Press, New 62. Murphy E. Social origins of depression in old age. Br York; 1986, pp. 165–73. J Psychiatry 1982; 141, 135–42. 77. Yesavage J. Geriatric Depression Scale. Psychophar- 63. Kessler RC. The effects of stressful life events on macol Bull. 1988; 24 (4): 709–10. depression. Annu Rev Psychol. 1997; 48: 191–214. 78. Wall J, Lichtenberg P, MacNeill S, Deshpande S, 64. Wikblad KF, Wibell LB, Montin KR. Health and Walsh P. Depression detection in geriatric rehabilita- unhealth in chronic disease. Scand J Caring Sci. 1991; tion: Geriatric Depression Scale Short Form vs. Long 5 (2): 71–7. Form. Clin Gerontol. 1999; 20 (1): 13–21. 400 CH 28 DEPRESSION AND DIABETES IN OLDER ADULTS

79. Stiles P, McGarrahan J. The Geriatric Depression 92. Wu JH, Haan MN, Liang J, Ghosh G, Gonzalez HM, Scale: a comprehensive review. J Clin Geropsychol. Herman WH. Diabetes as a predictor of change in 1998; 4 (1), 89. functional status among older Mexican Americans: A 80. Watson L, Pignone M. Screening accuracy for late-life population-based cohort study. Diabetes Care 2003; depression in primary care: A systematic review. J 26 (2): 314–19. Fam Pract 2003; 52 (12), 956–64. 81. Abas M, Phillip C, Carter J, Walter S, Banerjee 93. Black SA, Goodwin JS, Markides KS. The association S, Levy R. Culturally sensitive validation of between chronic diseases and depressive symptoma- screening questionnaires for depression in older tology in older Mexican Americans. J Gerontol. 1998; African-Caribbean people living in South London. 53A: M188–94. Br J Psychiatry. 1998; 173; 249–54. 94. Papassotiropoulos A, Heun R, Maier W. The impact 82. Rait G, Burns A, Baldwin R, Morley M, of dementia on the detection of depression in elderly Chew-Graham C, St Leger AS, et al. Screening for subjects from the general population. Psychol Med. depression in African Caribbean elders. Fam Pract. 1999; 29 (1), 113–20. 1999; 16 (6): 591–5. 83. Alexopoulos G, Abrams R, Young R, Shamoian C. 95. UK National Institute for Health and Clinical Excel- Cornell Scale for depression in dementia. Biol Psy- lence (NICE) Depression. Management of depression chiatry. 1988; 23 (3): 271–84. in Primary and Secondary Care. Clinical Guideline 84. Kørner A, Lauritzen L, Abelskov K, Gulmann N, 23 (amended). NICE, London; 2007. Brodersen AM, Wedervang-Jensen T, et al.TheGeri- 96. Jacobson AM, Weinger K. Treating depression in di- atric Depression Scale and the Cornell Scale for abetic patients: is there an alternative to medications? Depression in Dementia. A validity study. Nord J Psy- Ann Intern Med. 1998; 129 (8): 656–7. chiatry 2006; 60 (5): 360–4. 97. Alexopoulos G, Katz I, Reynolds C, Carpenter D, 85. Kurlowicz L, Evans L, Strumpf N, Maislin G. A Psy- chometric Evaluation of the Cornell Scale for Depres- Docherty JP. Expert consensus panel for pharma- sion in dementia in a frail, nursing home population. cotherapy of depressive disorders in older patients. Am J Geriatr Psychiatry 2002; 10 (5): 600–8. The expert consensus guideline series. Pharmacother- 86. Radloff LS. The CES-D scale: A self-report depres- apy of depressive disorders in older patients. Postgrad sion scale for research in the general population. Appl Med. 2001; Special No. Pharmacotherapy: 1–86. Psychol Measurement. 1977; 1: 385–401. 98. Wilson K, Mottram P, Sivanranthan A, Nightingale A. 87. Beck AT, Ward, CH, Mendelsohn M, Mock J, Er- Antidepressant versus placebo for depressed elderly. baugh J. An inventory for measuring depression, Arch Cochrane Database of Systematic Reviews 2001; (2): Gen Psychiatry 1961; 4: 561–71. 88. Lustman PJ, Clouse RE, Griffith LS, Carney RM CD000561. and Freedland KE. Screening for depression in dia- 99. Mottram P, Wilson K, Strobl J. Antidepressants for betes using the Beck Depression Inventory. Psycho- depressed elderly. Cochrane Database of Systematic som Med. 1997; 59 (1), 24–31. Reviews 2006; (1): CD003491. 89. Lustman PJ, Clouse RE, Griffith LS, Carney RM, 100. Gill D, Hatcher S. Antidepressants for depression Freedland KE. Screening for depression in diabetes in medical illness. Cochrane Database of Systematic using the Beck Depression Inventory. Psychosom Reviews 2000; (4): CD001312. Med. 1997; 59 (1): 24–31. 90. Hermanns N, Kulzer B, Krichbaum M, Kubiak T, 101. Gill, D. Hatcher, S. Antidepressants for depression Haak T. How to screen for depression and emotional in medical illness: Update. Cochrane Database of problems in patients with diabetes: comparison of Systematic Reviews 2007; (4): CD001312. screening characteristics of depression questionnaires, 102. Lustman PJ, Griffith LS, Clouse RE, Freedland KA, measurement of diabetes-specific emotional problems Eisen SA, Rubin EH, et al. Effects of nortriptyline on and standard clinical assessment. Diabetologia 2006; depression and glycemic control in diabetes: results 49 (3): 469–77. of a double-blind, placebo controlled trial. Psychosom 91. de Rekeneire N, Rooks RN, Simonsick EM, Shorr RI, Med. 1997; 59 (3): 241–50. Kuller LH, Schwartz AV, et al.Health,Agingand Body Composition Study. Diabetes is associated with 103. McIntyre R, Soczynska J, Konarski J, Kennedy S. The subclinical functional limitation in nondisabled older effect of antidepressants on glucose homeostasis and individuals: the Health, Aging, and Body Composition insulin sensitivity: synthesis and mechanisms. Expert study. Diabetes Care 2003; 26 (7): 3257–63. Opin Drug Saf. 2006; 5 (4): 157–68. 28.9 SUMMARY 401

104. Taylor D, Paton C, Kerwin R. The South London 118. UK National Institute for Health and Clinical Excel- and Maudsley & Oxleas Foundation Trust Prescribing lence (NICE) Depression. Management of depression Guidelines, 9th edition. Informa Healthcare, London; in Primary and Secondary Care. Clinical Guideline 2007. 23 (amended). NICE, London; 2007. 105. Bazire S. Psychotropic Drug Directory. Fivepin Pub- 119. Williams MM, Clouse RE, Nix BD, Rubin EH, Sayuk lishing, Wiltshire; 2003. GS, McGill JB, et al. Efficacy of sertraline in preven- 106. Lustman PJ, Freedland KE, Griffith LS, Clouse RE. tion of depression recurrence in older versus younger Fluoxetine for depression in diabetes: a randomized adults with diabetes. Diabetes Care 2007; 30 (4): double-blind placebo-controlled trial. Diabetes Care 801–6. 2000; 23 (5): 618–23. 120. Valenstein M, Vijan S, Zeber JE, Boehm K, Buttar A. 107. Gulseren L, GulserenS,HekimsoyZ,MeteL.Com-¨ The cost-utility of screening for depression in primary parison of fluoxetine and paroxetine in type II dia- care. Ann Intern Med. 2001; 134 (5): 345–60. betes mellitus patients. Arch Med Res. 2005; 36 (2): 121. Lustman PJ, Griffith LS, Freedland KE, Kissel SS, 159–65. Clouse RE. Cognitive behavior therapy for depres- 108. Goodnick P. Use of antidepressants in treatment of sion in type 2 diabetes. A randomized controlled comorbid diabetes mellitus and depression as well as trial. Ann Intern Med. 1998; 129 (8): 613–21. in diabetic neuropathy. Ann Clin Psychiatry 2001; 13 122. Lustman PJ, Clouse RE, Freedland KE. Management (1): 31–41. of major depression in adults with diabetes: implica- 109. Lustman PJ, Clouse RE, Nix BD, Freedland KE, Ru- tions of recent clinical trials. Sem Clin Neuropsychi- bin EH, McGill JB, et al. Sertraline for prevention of atry 1998; 3 (2): 102–14. depression recurrence in diabetes mellitus: a random- 123. UK National Institute for Health and Clinical Excel- ized, double-blind, placebo-controlled trial. Arch Gen lence (NICE) Depression. Management of depression Psychiatry 2006; 63 (5): 521–9. in Primary and Secondary Care. Clinical Guideline 110. Gray D, Fujioka K, Devine W, Bray G. A randomized 23 (amended). NICE, London; 2007. double-blind clinical trial of fluoxetine in obese dia- 124. Williams JW Jr, Katon W, Lin EH, Noel¨ PH, Worchel betics. Int J Obes Relat Metab Dis. 1992; 16 (Suppl. J, Cornell J, et al. The effectiveness of depression care 4): S67–72. management on diabetes-related outcomes in older 111. Maheux P, Ducros F, Bourque J, Garon J, Chiasson patients. Ann Intern Med. 2004; 140 (12): 1015–24. J-L. Fluoxetine improves insulin sensitivity in obese 125. Katon W, Von Korff M, Lin E, Simon G, Ludman patients with non-insulin-dependent diabetes mellitus E, Russo J, et al. The Pathways Study: a randomized independently of weight loss. Int J Obes Relat Metab trial of collaborative care in patients with diabetes Disord. 1997; 21 (2): 97–102. and depression. Arch Gen Psychiatry 2004; 61 (10): 112. Andrew, J.E., Prescott, P., Smith, T.M.F., Inman, 1042–9. W.H.W. and Kubota, K. Fluoxetine: another new 126. Huang X, Song L, Li T. The effect of social support on antidepressive. Drug and Therapeutics Bulletin 1990; type II diabetes with depression. Chin J Clin Psychol. 28, 33. 2001; 9, 187–9. 113. Bazire S. Psychotropic Drug Directory. 7th edition. 127. Ismail K, Winkley S, Rabe-Hesketh. Systematic re- Fivepin Publishing, Wiltshire; 2007. view and meta-analysis of randomised controlled 114. Raskin J, Smith T, Wong K, Pritchett Y, D’Souza trials of psychological interventions to improve gly- D, Iyengar S, et al. Duloxetine versus routine care caemic control in patients with type 2 diabetes. Lancet in the long-term management of diabetic peripheral 2004; 363 (9421): 1589–97. neuropathic pain. J Palliat Med. 2006; 9 (1): 29–40. 128. Cuijpers P, Smit F. Subthreshold depression as a risk 115. McIntyre R, Mancini D, Pearce M, Silverstone P, indicator for major depressive disorder: a systematic Chue P, Misener VL et al. Mood and psychotic review of prospective studies. Acta Psychiatr Scand. disorders and Type 2 diabetes: a metabolic triad. Can 2004; 109 (5): 325–31. J Diabet. 2005; 29, 122–32. 129. Paile-Hyvarinen M, Wahlbeck K, Eriksson JG. Qual- 116. Himmerich H, Fulda S, Schaaf L, Beitinger P, Schuld ity of life and metabolic status in mildly depressed A, Pollmacher¨ T. Changes in weight and glucose women with type 2 diabetes treated with paroxetine: tolerance during treatment with mirtazapine. Diabetes A single-blind randomised placebo controlled trial. Care 2006; 29 (1): 170. BMC Fam Pract. 2003; 4, 7. 117. Musselman D, Betan E, Larsen H, Phillips L. Re- 130. Christensen H, Griffiths KM, Jorm AF. Delivering lationship of depression to diabetes types 1 and 2: intervention for depression by using the internet: ran- epidemiology, biology, and treatment. Biol Psychiatry domised controlled trial. Br Med J. 2004; 328: 625. 2003; 54 (3): 317–29. 29 Diabetes and Falls

Christine T. Cigolle 1 and Caroline S. Blaum 2 1Department of Family Medicine, University of Michigan, Ann Arbor and the Ann Arbor VA Healthcare System Geriatric Research, Education and Clinical Center, Ann Arbor, MI, USA 2Department of Internal Medicine, University of Michigan, Ann Arbor and the Ann Arbor VA Healthcare System Geriatric Research, Education and Clinical Center, Ann Arbor, MI, USA

Diabetes has its greatest incidence and prevalence in Key messages the older adult population, and diabetes and its treatments, complications and comorbid conditions affect • Falling may be the presenting sign of diabetes the functioning of nearly all organ systems in the body. and other serious medical problems, and may Falling, likewise, has its greatest incidence and preva- result in considerable morbidity, disability and lence among older adults, and nearly all organ systems mortality. can be implicated in the pathophysiology underlying • Diabetes, via a number of different mechanisms falls. Thus, it is not surprising that the relationship be- including hyperglycaemia, hypoglycaemia, sen- tween diabetes and falling is interconnected and com- sory impairment, medication use, as well as balance, mobility and gait disorders, may cause plex. Compared to those without the disease, older a substantial increase in the risk of falls. adults with diabetes are at increased risk for falling, • All older adults with diabetes should be screened with diabetes contributing multiple causal factors for annually for falls. falls.

29.1 Introduction 29.2 Deﬁnition

Falling has been long recognized as a geriatric condi- The deﬁnition of a fall includes several criteria. First, there is a change in body position, such that the indi- tion (or geriatric syndrome). It is common in the older vidual comes to rest at a lower level, not necessarily on adult population and is generally multifactorial in ae- the ﬂoor or ground. Next, the change in body position tiology. Falling may be the presenting sign of serious is unintentional or inadvertent. Last, most authorities medical problems, and falls themselves result in con- agree that a fall occurs without loss of consciousness siderable morbidity, disability and mortality. They are (in contrast to syncope and seizures) or trauma (such as also responsible for substantial economic costs. For a motor vehicle-pedestrian accident). Physically, a fall example, health care expenditures for fall-related in- results when the individual’s centre of gravity moves beyond his or her base of support, and there is inade- juries in adults aged ≥65 years totalled more than $US quate effort to maintain or restore balance [3]. 19 billion in the United States in 2000 [1, 2].

Diabetes in Old Age Third Edition Edited by Alan J. Sinclair © 2009 John Wiley & Sons Ltd. ISBN: 978-0-470-06562-4 404 CH 29 DIABETES AND FALLS Falls are frequently described with reference to their the disease. Of older adults with diabetes, falling is frequency, their outcome, and their association with more common among those who have diabetes com- balance or movement disorders. Falls in older adults plications [6]. Among older adults with chronic kidney are most concerning when the individual has multiple disease requiring dialysis (including but not limited to falls, has falls resulting in injury, or has an associated those with diabetes), the incidence of falls has been balance or gait disorder. Falling is often accompanied estimated to be 1.2 to 1.6 per patient-year (pt-yr) [7–9]. by an associated condition, a fear of falling. Here, an In the community, more than 50% of falls occur in the home. Approximately 15% of falls in the hospital individual self-limits his activity due to a fear of and setting lead to serious injury, and are associated with a desire to avoid falling and fall-related injury. a longer length of stay and increased admission to long-stay nursing facilities. Fall rates are higher for 29.3 Epidemiology those residing in long-stay nursing facilities, with an incidence rate of over 1% per pt-yr [10]. 29.3.1 Falls 29.3.2 Complications of falls The prevalence of falling in the older adult popula- Serious complications of falls include death, injury, ≥ tion is approximately 30% for adults aged 65 years a long period of being down and unable to get up [4]. Falling may be more prevalent among older adults following a fall, loss of independence, and fear of having diabetes compared to those without the dis- falling. ease. Data from the Health and Retirement Study, a Fractures and other serious injuries are a feared nationally representative, longitudinal health interview outcome of falling. Some 10% of falls in the older survey, have revealed falls to be more common across adult population result in non-fatal injury. Falling is the age spectrum for older adults with diabetes, as the most common cause of hospital admission for compared to those without diabetes (Figure 29.1) [5]. trauma in the older adult population, with falls being Falling has similarly been shown to be more prevalent responsible for 87% of all fractures and over 95% of in older women with diabetes than in those not having all hip fractures [11].

50 45 40 35 30 25 20

Prevalence (%) 15 10 5 0 65-69 70-74 75-79 80-84 85-89 >=90 Age (years)

Diabetes, >=2 Falls No diabetes, >=2 Falls Diabetes, >=1 Injurious Falls No diabetes, >=1 Injurious Falls

Figure 29.1 Prevalence of falls by diabetes status and age. Weighted percentages derived using the Health and Retirement Study (HRS) respondent population weights to adjust for the complex sampling design of the HRS survey (these data are from the 2004 wave of the HRS). 29.4 PATHOPHYSIOLOGY OF, AND RISK FACTORS FOR, FALLING 405

A long period of being immobile on the floor or Table 29.1 Risk factors for falling in older adults with ground after a fall and prior to discovery can result in diabetes. dehydration, rhabdomyolysis and aspiration pneumo- • Hyperglycaemia, hypoglycaemia nia. The loss of independence – including the need for • Multisensory impairment: vision, hearing, vestibular, placement in a long-stay nursing facility – is another proprioceptive feared outcome for older adults. Indeed, over 50% of • Dehydration, fluid overload/oedema older adults who have a hip fracture are unable to live • Postural hypotension independently thereafter. • Neuropathy, autonomic and peripheral The fear of falling, and attempts to decrease the likelihood of falling, can lead to a cascade of events: • Foot disorders restriction of activities, including both social participa- • Balance, mobility, and gait disorders tion and physical activities; deconditioning; decline in • Cognitive impairment functional status; and, finally, placement in a long-stay • nursing facility. Depression Older adults with diabetes are at risk for worse • Comorbid diseases and conditions: cardiovascular, cere- consequences from falls, due to a variety of causes, brovascular, peripheral vascular, anaemia, incontinence, including an increased risk of fracture. sleep disorders, arthritis, etc. • Medications, especially centrally acting and psychoactive 29.4 Pathophysiology of, and risk drugs and cardiovascular agents • factors for, falling Polypharmacy

29.4.1 Multifactorial in Aetiology 29.4.3 Hyperglycaemia and hypoglycaemia As a geriatric condition or geriatric syndrome, falling typically has a complex aetiology; that is, falling is the Hyperglycaemia can cause vision impairment, dizzi- result of the “...accumulated effects of impairments ness, increased urination due to osmotic diuresis, and in multiple systems” [12]. Risk factors for falling may dehydration, each of which can lead to falls. Hypogly- be classified as intrinsic versus extrinsic [13]: caemia can cause confusion and imbalance, also resulting in falls. Lower haemoglobin A levels, reflecting • 1c Intrinsic, or personal, causes include age-related lower blood glucose levels, have been associated with changes, chronic diseases, acute illnesses, medica- falling in frail older adults [15]. Older adults with dia- tions and mobility. betes may have a decreased ability to perceive the onset • Extrinsic causes are those in the environment and and occurrence of both hyperglycaemia and hypogly- include items such as room design, lighting, walking caemia, and thus fail to take appropriate corrective surfaces, stairs, and obstructions and impediments. actions. There are no randomized controlled trials indicating the degree of glycaemic control that is most 29.4.2 The role of diabetes appropriate and beneficial for vulnerable older adults, Research studies continue to disclose the complex re- nor is there any evidence for the level of glycaemic lationship between diabetes and falling. Diabetes as a control that decreases the risk of falling. disease – and its management, complications and comorbid conditions – interacts with ongoing age-related 29.4.4 Sensory impairment functional decline in leading to falls in the older adult population [14]. As diabetes progresses in the older Sensory impairment contributes to falling by causing adult, diabetes complications and impairments accu- older adults to become disoriented in their physical en- mulate in multiple systems. Diabetes becomes a major vironment. Diabetes is directly responsible for several contributor to – and interacts with – many of the risk serious sensory impairments, which interact with and factors for falling, and so plays a substantial and multi- exacerbate other sensory impairments common in older faceted role in the pathophysiology underlying falling adults resulting from physiologic changes of aging and (Table 29.1). comorbid conditions. 406 CH 29 DIABETES AND FALLS

Vision congestive heart failure) and medications (e.g. diuretics). Hyperglycaemia resulting from poorly Eye diseases prevalent in the older adult population controlled diabetes can cause an osmotic diuresis include refractive error (e.g. far-sightedness), cataracts, leading to dehydration. Both, diabetic nephropathy glaucoma and macular degeneration. Diabetes causes and its treatment affect the fluid status of older adults, diabetic retinopathy and is a risk factor for cataracts thereby increasing the complexity of their medication and glaucoma. regimens and predisposing them to falls. Predialysis Impairment in vision in older adults is associated systolic blood pressure levels have been correlated with limitations in mobility, in activities of daily liv- with falls in dialysis patients. Other associations with ing, and in the instrumental activities of daily living. falls in the dialysis population include advancing The impairments in vision which most likely to lead age, increased comorbidity, medications (especially to falls are those that most directly affect everyday antidepressants), a previous history of falling and the functioning. Visual field loss, as compared to declines use of assistive devices [8, 9]. in visual acuity, stereoacuity and contrast sensitivity, is the component of vision most associated with an increased risk for falls [16–19]. 29.4.6 Postural hypotension Vision difficulties can be a factor in patient medication errors, especially in the setting of polypharmacy, Diabetes contributes to, or is associated with, multiple and so also indirectly contribute to falling. Older adults aetiologies of postural instability including: needing to read glucometer results and fill insulin sy- • Age-related physiological changes, such as diastolic ringes are at particular risk. dysfunction, vascular stiffness and decreased renal salt and water conservation. Hearing • Postprandial and exercise-induced hypotension. Hearing impairment is common in older adults, and diabetes has been shown to be associated with early de- • Autonomic insufficiency, both central and periph- velopment of sensorineural hearing loss [20]. Hearing eral. impairment, by distorting environmental input about an • Cardiac disease, including myocardial ischaemia and individual’s position in space, may contribute to falls. arrhythmia. It may be synergistic with vision impairment in causing functional decline. • Infection. • Dehydration. Vestibular • A loss of vestibular function leads to disorders of Medications, including alpha-receptor blockers, di- balance, especially with changes in the body position. uretics, beta-receptor blockers, vasodilators and other antihypertensives, opiates and sedative-hypnotics, Proprioceptive antipsychotics and phosphodiesterase-5 inhibitors. Proprioceptive impairment, with diabetic neuropathy being a significant cause (see below), leads to disori- 29.4.7 Neuropathy entation with ambulation. Older adults are at greatest risk for falling in those circumstances when proprio- • Autonomic: Diabetic autonomic neuropathy can ceptive input is most relied upon, for example, when cause hypotension, particularly postural hypotension walking on uneven surfaces or in dark or poorly lit and diarrhoea, and may have a role in muscle areas. weakness. • Peripheral: Diabetic peripheral neuropathy is a sig- 29.4.5 Alterations in fluid status nificant cause of balance and gait disorders and of Older adults with diabetes are particularly vulnerable falls in the older adult population. Motor neuropathy to alterations in fluid status, which may be due to a contributes to formation of calluses, hammer-toes variety of causes, including physiological changes of and other deformities of the feet. Sensory neuropa- aging (e.g. decreased thirst), comorbid diseases (e.g. thy causes loss of sensation, paraethesias and chronic 29.4 PATHOPHYSIOLOGY OF, AND RISK FACTORS FOR, FALLING 407 pain, and can ultimately lead to a lack of coordina- retinopathy, nephropathy and neuropathy. Macrovas- tion of gait and to the development and poor healing cular complications causing falls include coronary of foot ulcers. artery disease (myocardial ischaemia, congestive heart failure, arrhythmia), cerebrovascular disease (transient 29.4.8 Foot disorders ischaemic attacks, strokes) and peripheral vascular disease (claudication, foot ulceration, amputation). Disorders of the feet and ankles, resulting from the The comorbid condition most associated with falling neuropathic and vascular complications of diabetes, in- in older adults with diabetes is previous stroke. clude loss of sensation, pain, wounds and amputation. Anaemia of chronic disease, anaemia resulting from These disorders can lead to a reduced walking speed, chronic kidney disease and anaemia associated with the balance difficulties and falling. Toe strength plays a use of angiotensin-converting enzyme (ACE) inhibitors critical role in maintaining balance when the body’s (in susceptible individuals) may complicate diabetes in centre of gravity is displaced, and weakness or ampu- older adults. tation decreases the body’s ability to avoid or break a Urinary incontinence, the need for frequent toilet- fall [21, 22]. ing (due to hyperglycaemia or diuretic use) and dependency in toileting all can be associated with falls. 29.4.9 Balance, mobility and gait disorders Diabetes is a leading aetiology for neurogenic bladder Older adults with diabetes have reduced self-reported and overflow incontinence. Urge incontinence precipi- physical functioning, and also have a decreased per- tates the need for rapid access to toileting; functional formance on standard physical performance measures incontinence occurs when the older adult is unable to of balance and mobility [23, 24]. Vision, hearing and toilet in an expeditious manner. In trying to toilet to vestibular impairments; changes in fluid status; postu- avoid an accident, older adults can increase their risk ral instability; neuropathy; and foot problems all con- for falling. tribute to disorders in balance, mobility and gait in this population. These disorders interact with musculoskeletal diseases such as osteoarthritis to contribute 29.4.12 Other diseases and conditions to: (i) difficulty and dependency in transferring and in Diseases not necessarily related to diabetes that in- mobility functions; and (ii) falling. crease fall risk include sleep disorders (associated with obesity and the metabolic syndrome), restless leg syn- 29.4.10 Cognitive impairment and drome, other autonomic and peripheral neuropathies depression and arthritis disorders. Diabetes is postulated to have an aetiology in Alzheimer’s disease as well as vascular dementia. 29.4.13 Overall health status Diabetes is also strongly associated with depression, and both cognitive impairment and depression can lead Older adults who report fair or poor health or a decline to impairments in judgement and problem-solving. in health over the previous year have been shown to Depression causes impairment in concentration, while be more likely to fall [25]. advancing dementia causes apraxia, and impairments in these executive functions predispose to falling. Impairment in immediate recall may be especially 29.4.14 Medications predictive of fall risk. Both severity of cognitive Multiple medications – either individually or in com- impairment and change in cognitive function (i.e. a bination with other medications and conditions – recent decline in cognitive status) are associated with predispose older adults to falling. Central nervous sys- falling. tem and cardiovascular drug classes are particularly problematic [26, 27]. 29.4.11 Diabetes complications and At-risk medications include the following: comorbid diseases and conditions • opiates Multiple diabetes complications and comorbid diseases and conditions are associated with falling. Microvas- • sedative-hypnotics and anxiolytics (especially ben- cular complications causing falls include diabetic zodiazepines) 408 CH 29 DIABETES AND FALLS • antidepressants (especially tricyclic antidepressants; falls for those taking more than four different drugs selective serotonin reuptake inhibitors are associated concurrently [29, 30]. with falls to a lesser extent) • anticonvulsants 29.5 Risk factors for fall injury • anticholinesterase inhibitors Injury resulting from a fall is a function of the force • antipsychotics of the fall, and of the part of the body impacted in • antiparkinsonian agents the fall. Falls from greater height (e.g. a stairway or a bed using bedrails in the upright position), an inability • muscle relaxants to break a fall (e.g. an inability to reach for and hold • diuretics onto a handrail or piece of furniture) and landing on a hard surface all increase the force of a fall. Older adults • vasodilators have slowed reaction times and impaired protective re- • alpha-blockers sponses, and so may be unable to avoid injury as a con- • beta-blockers sequence of falling. The physical impact of falls may be increased in older adults having a lower body mass • other antihypertensives index, and who have decreased musculature and sub- • anti-arrhythmics (especially class 1A anti-arrhy- cutaneous tissues. Fracture is more likely with certain thmics) types of fall; for example, falling sideways on the fe- mur is more likely to lead to hip fracture, while falling • diabetes medications (especially those with pro- on an outstretched hand may lead to wrist fractures. nounced hypoglycaemic effects such as insulin and Older adults with diabetes may have worse conse- the sulphonylureas) quences from falls, due not only to slow wound healing • medications to treat erectile dysfunction (e.g. that can delay recovery but also to an increased risk of phosphodiesterase-5 inhibitors), and fractures with falling. Diabetes is a risk factor for osteoporosis and for osteoporotic fractures resulting from • all medications with anticholinergic effects. falls. The Study of Osteoporotic Fractures found that The antipsychotics are well known for their risk postmenopausal women with diabetes had an increased for falls, especially for older adults in institutional risk for falling and for hip and proximal humerus frac- settings. This drug class is also problematic for its tures with falling [31, 32]. The Women’s Health Initia- effects on glucose and lipid metabolism, worsening the tive found that postmenopausal women with diabetes metabolic syndrome and so exacerbating both diabetes had an increased risk for fractures overall and increased and vascular disease. risks of hip, foot and spine fractures separately, despite Findings from the United Kingdom Prospective Di- having a higher bone mineral density [33]. These find- abetes Study showed that a tight control of blood glu- ings may reflect differences in the structure of bone. cose levels was associated with twice as many episodes Older adults with diabetes frequently have comorbid of hypoglycaemia. Medications with pronounced hy- cardiac and vascular diseases warranting anticoagula- poglycaemic effects, such as insulin and the sulphony- tion which, itself, can lead to an increase in injuries as lureas, can be especially challenging in this regard [14, a result of falling. 15, 28]. Combinations of certain medications increase the risk of fall for older adults (e.g. multiple agents hav- 29.6 Evaluation of older adults with ing psychoactive effects or anticholinergic side effects). diabetes who fall Certain drug-drug interactions (e.g. when one agent is a benzodiazepine) result from alterations in hep- 29.6.1 Screening atic metabolism or renal excretion, perhaps as a consequence of diabetic nephropathy. Polypharmacy is Older adults with diabetes should be screened yearly well-documented as a risk for falling, with research for falls. Those having two or more falls in the previ- studies having demonstrated an increased number of ous year, a single fall and an associated gait or balance 29.6 EVALUATION OF OLDER ADULTS WITH DIABETES WHO FALL 409 disorder, or an injurious fall should be further eval- • Details of current medications, including over- uated. The aims of screening and evaluation are to the-counter products. Recent changes in the med- identify potentially reversible or modifiable causes and ication regimen, including any new medications, conditions for falling, and to develop a management should be investigated. The use of high-risk plan to limit falls and fall injuries. The evaluation of medications should prompt questioning about the older adults with diabetes who fall is not substantially presence of any side effects and their severity. different from the evaluation of those without diabetes Similarly, the use of anticoagulants should prompt who fall. Rather, older adults with diabetes can have questioning about compliance, bleeding episodes, a greater burden of comorbid conditions predisposing etc. An overall assessment of polypharmacy should to falls and a greater potential for worse consequences be made. from falling [34–36]. • The pattern of alcohol use. 29.6.2 History • The patient’s functional status, including impair- Pertinent history includes the circumstances surround- ments in activities of daily living (ADLs), instrumen- ing the fall(s) and associated medical and social con- tal activities of daily living (IADLs) and mobility. ditions, and should be obtained from the patient and, • Details of recent transitions which can be associated if appropriate and possible, from caregivers. Older with an increased risk for falling (e.g. acute illness, adults with diabetes should be questioned specifically discharge from a hospital or nursing facility, new about glycaemic management and diabetes complica- medications). tions. History-taking should cover: • The availability and level of family and social sup- • The circumstances of the fall(s), including location, port. time of day, activity associated with the fall, etc. • Associated symptoms, including presyncope (and syncope), light-headedness, dizziness, confusion, 29.6.3 Physical examination palpitations, hypoglycaemic symptoms, etc. The physical examination should be directed by the • Any history of previous falls, including their fre- patient’s history, and include targeted assessment of quency and pattern. cardiovascular, neurological and musculoskeletal systems. The examination should include a search for • Details of high-risk behaviours, such as certain home physical findings that indicate the state of diabetes and repairs, outdoor activities, etc. its complications: • Environmental hazards [37]. • Vital signs, including orthostatic pulse and blood • The status of the patient’s diabetes and its manage- pressure. Vital signs should be measured, first, after ment, including frequency and degree of hypergly- the patient has been supine for 5 min; next, imme- caemia and hypoglycaemia, current management and diately upon standing; and last, after standing for interventions, diabetes complications, etc. 2 min. A 20 mmHg drop in systolic pressure in either standing position indicates orthostatic hypotension • The presence of comorbid diseases and conditions, [36]. including heart disease (coronary artery disease, congestive heart failure, arrhythmia), anaemia, • Head and neck examination, including assessment seizure disorder, stroke, Parkinson’s disease and of vision, hearing, the thyroid gland and carotid other movement disorders, cognitive impairment, arteries. neuropathy, depression/anxiety, sensory impairment, • Cardiac examination, including evaluation for ar- arthritis, etc. rhythmia and assessment of volume status (dehydra- • The patient’s overall cognitive status and any recent tion, oedema). change in cognition (e.g. delirium). • Neurological examination to assess cognitive status, • The presence of other geriatric conditions and symp- sensory function (proprioception), muscle strength, toms (e.g. incontinence, nocturia) and their relation cerebellar function, tremor, as well as balance, to falls. movement and gait difficulties. The evaluation of 410 CH 29 DIABETES AND FALLS sensory function should include the extent and sever- fall-related injuries, to maintain independence, and to ity of sensory loss due to diabetic neuropathy. A address the fear of falling [34–36]. timed ‘Get-Up and Go’ test may be considered to Management strategies for this population come evaluate mobility and the ability to stand from a with a number of inherent challenges. Because falls seated position [38, 39]. Older adults using assistive have multiple possible causes and, therefore, multi- devices should be evaluated for their correct use. ple possible interventions (Table 29.2), effective fall management requires targeting interventions to indi- • Musculoskeletal examination to assess joints for vidual patients and their environments. The typical range of motion and presence of arthritic changes. older adult with diabetes has a multifactorial aetiology • Foot examination, including the evaluation of ulcer- for falling, necessitating multicomponent intervention ations, deformities and painful lesions. strategies. For each patient, the benefits of each intervention must be weighed against its burdens and risks. 29.6.4 Laboratory evaluation and Many interventions are time- and effort-demanding for radiological imaging older adults and their caregivers. Some interventions require ongoing monitoring and optimization (e.g. gly- Laboratory testing and imaging should be directed by caemic and hypertension medication regimens, correct the results of the history and physical examination. use of assistive devices). The benefits of some inter- Glucose monitoring and HbA level are important 1c ventions can be time-limited in nature (e.g. physical as indicators of glycaemic control in diabetes man- therapy). Interventions frequently entail the physician agement. The following may also be considered, de- coordinating care with other providers, such as physical pending on the particularities of the fall syndrome: therapists, occupational therapists and social workers. complete blood count; electrolytes; renal function; Older adults with diabetes who fall almost always have thyroid-stimulating hormone; vitamin B ; electrocar- 12 multiple comorbid diseases and conditions; fall man- diogram; brain imaging; and bone density scan. Indica- agement interventions can conflict with evidence-based tions for imaging of the brain include a history of head guidelines for diabetes and those comorbidities [40]. injury and focal deficits on neurological examination. Finally, in some health care systems and settings, 29.6.5 Assessment of older adults in physician reimbursement may be geared toward diag- hospital or nursing facility settings nosis and the management of diseases and not toward prevention – which is the primary focus of fall man- Older adults with diabetes are more likely to be hospi- agement interventions. talized and to reside in long-stay nursing facilities than those without diabetes, and represent ‘special’ pop- 29.7.1 Glycaemic status ulations when evaluating falls. These patients should be initially assessed and then regularly monitored for It appears that no conclusive studies have been con- acute cognitive changes (e.g. delirium) which might in- ducted to determine the optimal level of glycaemic crease fall risk during hospitalization. Similarly, older control for the frail or vulnerable older adult popula- adults with diabetes in institutional settings are at risk tion; neither is any evidence-derived algorithm avail- for nosocomially acquired dehydration and infection able to stratify older adults for glycaemic control based (e.g. pneumonia, urinary tract infection), and so should on their medical and functional status. Falling is one be monitored for these conditions, which contribute important consideration in the context of the overall to falls if not diagnosed. Older adults with difficulty health of individuals in deciding on targets for diabetes or dependency in transferring and toileting are at risk management. For older adults with diabetes who fall, for falls when attempting these tasks, especially when the optimal level of glycaemic control should be de- without assistance. termined on an individual basis. Those with recurrent falls especially need to avoid hypoglycaemia. Medica- tions with potent hypoglycaemic effects (e.g. insulin, 29.7 Management of older adults sulphonylureas) should be used cautiously, particularly with diabetes who fall in those older adults who are especially vulnerable, who live alone, who have cognitive impairment, or The goals of management for older adults with dia- who have previously demonstrated inconsistent eating betes who fall are to prevent future falls, to prevent patterns (e.g. missing meals). 29.7 MANAGEMENT OF OLDER ADULTS WITH DIABETES WHO FALL 411

Table 29.2 Management interventions for older adults with diabetes who fall. • Avoid hyperglycaemia and hypoglycaemia. • Ophthalmology consultation. Treatment of diabetic retinopathy, cataracts, glaucoma, macular degeneration, etc. Corrective lenses. Environmental modifications. • Audiology consultation. Hearing aids and other assistive devices. • Address underlying causes of postural imbalance. Eliminate or decrease medication doses. Compression stockings. Liberalize fluid, salt, and caffeine restrictions. • Referral to podiatry. Appropriate footwear. • Address underlying causes of postural imbalance, movement disorders, and impairments in gait or balance. Referral to physical and occupational therapy. Strength/resistance training; range of motion exercises. Gait and balance training. Pos- tural awareness exercises. Appropriate assistive devices. • Regular assessment of diabetes comorbid conditions for progression and change in impact on mobility and risk for falls. Continual optimization of management of comorbidities. Treatment/management of depression and cognitive impairment. • Minimize medications that are centrally acting and/or that have psychoactive and anticholinergic side effects. Minimize polypharmacy. Consider vitamin D supplementation. • Review benefits and risks of anticoagulation. • Consider treatment of osteoporosis. • Home safety assessment and environmental interventions.

29.7.2 Sensory impairments 29.7.4 Neuropathy and foot disorders Vision and hearing impairments should be further as- Older adults with diabetes should be referred to podia- sessed (e.g. ophthalmology and audiology consulta- try when appropriate. Footwear that minimizes balance tions) and managed with the appropriate interven- and gait difficulties (e.g. low heel, thin sole) should be tions (e.g. cataract surgery, laser therapy for diabetic selected. retinopathy) and assistive devices (corrective lenses, hearing aids). Environmental modifications for older 29.7.5 Balance, mobility and gait disorders adults with diabetes who have vision impairment include lighting in the home and glare reduction. Mo- Older adults with movement disorders or impairments bility training for the visually impaired can also be in balance or gait can be referred for physical and/or considered. occupational therapy. Physical therapists can assess impaired arm or leg strength and/or range of motion and 29.7.3 Alterations in fluid balance and recommend strength/resistance training or range of motion exercises. Therapists can also provide balance and postural hypotension gait training. Older adults may benefit from postural Management interventions should be directed to cor- awareness exercises, such as tai chi or yoga. Appropri- recting or treating the underlying causes. Diuretics, ate assistive devices, such as canes and walkers, which vasodilators, alpha-blockers and other antihypertensive provide a wider base of support, should be prescribed medications should be used with caution (and at the and assessed for their correct use. lowest dose that provides benefit) in older adults who fall. Other interventions that may be considered include 29.7.6 Comorbid diseases and conditions use of compression stockings and the liberalization of fluid, salt and caffeine restrictions. Older adults and Diabetes comorbid diseases and conditions should be their caregivers should be educated about the risk of regularly evaluated for their progression and changing falling with postural changes. impact on mobility and risk for falls (e.g. progression 412 CH 29 DIABETES AND FALLS of coronary artery disease, increasing frequency of uri- anticoagulation in this population is the subject of on- nary incontinence). Management should be continually going research; physicians and other providers should optimized to lessen the risk for falls, with providers examine research study population samples and out- periodically screening for depression and cognitive im- comes to determine if these are applicable to their own pairment. Effective depression treatment can improve patient populations and settings. Older adults continu- attention and concentration deficits that can lead to ing on anticoagulants should be regularly reassessed falls, but many agents have side effects that can infor fall risk. crease fall risk. The management of cognitive impair- The treatment of osteoporosis can decrease the inci- ment includes minimizing offending medications and dence of fall-related fractures. However, medications providing a supportive environment. Older adults with currently indicated for osteoporosis can have prob- cognitive impairment should be prescribed assistive de- lematic side effects in older adults with diabetes: cal- vices appropriate to their cognitive ability. cium binds many medications, making dosing regimens difficult for older adults with polypharmacy; bisphos- 29.7.7 Medications phonates are contraindicated in individuals with upper gastrointestinal disease, such as diabetic gastroparesis; Medications that are centrally acting and have psy- and oestrogen increases the incidence of cardiovascu- choactive effects should be assessed for their overall lar and thromboembolic events and so is contraindi- benefit versus their risk for falls. Medications with anti- cated in those with known disease. Hip protectors – a cholinergic effects should be similarly evaluated. Any non-pharmacological intervention – have poor adher- medications implicated in postural hypotension or in ence rates, and the evidence for their ability to pre- balance, mobility or gait disorders should be discon- vent hip fractures is conflicting. One recent study with tinued, if possible. good adherence rates failed to show any benefit in Polypharmacy should be addressed, with the aim of a long-stay nursing facility population [44]. One fi- eliminating medications or reducing their dosage, espe- nal intervention to consider is routine weight-bearing cially those that are risk factors for falls. Consideration exercise. should be given to eliminating medications that have benefit but are of low priority in an individual patient’s 29.7.9 Environmental interventions treatment plans (e.g. ACE inhibitors or beta-blockers in patients with advanced cancer). The resulting med- Home safety assessment and recommended environ- ication regimens should be examined for drug–drug mental interventions should target each older adult’s interactions and for the optimal dosing of medications particular fall risks, with emphasis placed on the cir- (e.g. spreading antihypertensive medications through- cumstances of previous falls. Older adults in new or out the day, taking certain medications at bedtime). unfamiliar settings may require additional care and at- Finally, ongoing medication reconciliation is required, tention. with continual attention to optimal dosing, the poten- Lighting should be optimized, especially for older tial for medication interactions, and the correct use of adults with vision impairment. Lighting at night must medications by older patients. The results of several be adequate to enable safe night-time toileting. Con- studies have indicated that vitamin D supplementation trasts can be used to highlight steps and stairways; high decreases the risk for falls, most likely by its benefi- gloss and high glare flooring should be avoided. cial effects on muscle function [41, 42]. However, to Caregivers should attend to barriers at floor level, be effective and safe the total daily dosing should be including slippery floors, carpeting and throw rugs, 800–2000 IU and accompanied by an adequate calcium electric cords and obstructions, and impediments such supplementation [43]. as furniture and clutter. Older adults – especially those with impairments in mobility and transferring – often use the furniture as a support; consequently, any fur- 29.7.8 Fall-related injury niture should be stable and of the correct height and Older adults with certain cardiac and vascular diabetes without sharp edges. Railings and grab bars can be comorbidities may be candidates for anticoagulation. placed in the home to assist with mobility and to help A discussion with patients and their caregivers should the older adult avoid or break a fall. address the anticoagulation’s benefits and its risks for The bathroom is a high risk environment for falling, fall-related injury (e.g. intracranial bleeding). Of note, with bathroom floor mats a particular hazard. Older 29.8 PATIENT SAFETY AND QUALITY OF CARE 413 adults are at risk for falls when getting into and out of Commission on the Accreditation of Healthcare Orga- the bath and shower, and may be aided by the use of nizations (JCAHO) in the United States). The Assess- shower chairs and the judicious placement of grab bars. ing Care of the Vulnerable Elders (ACOVE) initiative Individuals with urge incontinence, frequent urination, developed the Vulnerable Elders Survey (VES-13), a nocturia or difficulty or dependency in toileting, may clinical measure of vulnerability whose goal was to also benefit from grab bars and night lighting. identify older adults at risk for functional decline or death [45]. The ACOVE initiative has since devel- 29.7.10 Hospital or nursing facility settings oped quality indicators for a number of diseases and conditions prevalent among vulnerable older adults, in- Older adults with diabetes in hospital and nursing fa- cluding diabetes and falls [46, 47]. The complexity cility settings have unique fall risks. Providers should underlying falling in the older adult population with avoid prescribing medications and other interventions diabetes presents a challenge for guidelines and quality in the hospital that can precipitate or worsen acute cog- indicators to improve outcomes at the population level nitive changes (e.g. delirium) and thereby increase the and yet remain flexible to accommodate the needs of risk for falling. Likewise, polypharmacy and the use of individual patients. psychotropic medications and medications that lead to dehydration should be minimized in long-stay nursing facilities. In both settings, it is important to coordinate the patient’s medication regimens with other parts References of their daily schedules. For example, those receiving hypoglycaemic agents need to receive their meals on 1. Centers for Disease Control and Prevention. (2008) Costs of Falls Among Older Adults. Available at: time; the dosing of antihypertensives should also take http://www.cdc.gov/ncipc/factsheets/fallcost.htm. into account the possibility of postprandial hypoten- 2. Stevens, J.A., Corso, P.S., Finkelstein, E.A. and Miller, sion. T.R. (2006) The costs of fatal and non-fatal falls among Physical restraints should be avoided. Bedrails in older adults. Inj Prev, 12, 290–5. the upright position may act as physical restraints and 3. King, M.B. (2003) Falls. In: Hazzard, W.R., Blass, J.P., lead to falls from a greater height than otherwise. Halter, J.B., Ouslander, J.G. and Tinetti, M.E. (eds) Low beds and bedside mattresses are alternatives that Principles of Geriatric Medicine and Gerontology. 5th decrease the risk of fall-related injury. Other preventive edition. McGraw-Hill, New York, pp. 1517–29. measures should focus on transfers and on toileting 4. Rubenstein, L.Z. and Josephson, K. R. (2002) The (e.g. frequent toileting for patients undergoing diuretic epidemiology of falls and syncope. Clin Geriatr Med, treatment). In general, caregivers should search for 18, 141–58. the least restrictive means by which older adults can 5. Cigolle, C. T., Lee, P. G., Tian, Z. and Blaum, C. S. maintain independence in a safe fashion (e.g. correct (2008) Geriatric conditions and diabetes: The Health use of assistive devices, assistance in transferring). and Retirement Study. Journal of the American Geri- One critical component to reducing falls and atrics Society, 56 (s1), s186–7. fall-related injuries in institutional settings is the 6. Wallace, C., Reiber, G. E., Lemaster, J., Smith, D. G., ongoing education of nursing staff and other care Sullivan, K., Hayes, S. and Vath, C. (2002) Incidence of falls, risk factors for falls, and fall-related fractures providers. in individuals with diabetes and a prior foot ulcer. Diabetes Care, 25, 1983–6. 29.8 Patient safety and quality of 7. Cook, W. L. and Jassal, S. V. (2005) Prevalence of falls among seniors maintained on hemodialysis. Int care Urol Nephrol, 37, 649–52. 8. Cook, W. L., Tomlinson, G., Donaldson, M., The impact of falls on morbidity, disability, loss of Markowitz, S. N., Naglie, G., Sobolev, B. and Jassal, independence and mortality in the older adult popula- S. V. (2006) Falls and fall-related injuries in older tion is recognized by health policy decision makers. dialysis patients. Clin J Am Soc Nephrol, 1, 1197–204. Falls and fall-related injuries are an ongoing focus of 9. Desmet, C., Beguin, C., Swine, C. and Jadoul, M. regulatory agencies, initiatives and guidelines aimed at (2005) Falls in hemodialysis patients: prospective study improving patient safety and the quality of care in hos- of incidence, risk factors, and complications. Am J pital, long-term care and outpatient settings (e.g. Joint Kidney Dis, 45, 148–53. 414 CH 29 DIABETES AND FALLS

10. Ray, W. A., Taylor, J. A., Brown, A. K., Gideon, 22. Menz, H. B., Morris, M. E. and Lord, S. R. (2006) P., Hall, K., Arbogast, P. and Meredith, S. (2005) Foot and ankle risk factors for falls in older people: a Prevention of fall-related injuries in long-term care: prospective study. J Gerontol A Biol Sci Med Sci, 61, a randomized controlled trial of staff education. Arch 866–70. Intern Med, 165, 2293–8. 23. Volpato, S., Blaum, C., Resnick, H., Ferrucci, L., Fried, 11. Fife, D. and Barancik, J. I. (1985) Northeastern Ohio L. P. and Guralnik, J. M. (2002) Comorbidities and im- Trauma Study III: incidence of fractures. Ann Emerg pairments explaining the association between diabetes Med, 14, 244–8. and lower extremity disability: The Women’s Health 12. Tinetti, M. E., Inouye, S. K., Gill, T. M. and Doucette, and Aging Study. Diabetes Care, 25, 678–83. J. T. (1995) Shared risk factors for falls, incontinence, 24. Volpato, S., Ferrucci, L., Blaum, C., Ostir, G., Cap- and functional dependence. Unifying the approach to pola, A., Fried, L. P., Fellin, R. and Guralnik, J. geriatric syndromes. JAMA, 273, 1348–53. M. (2003) Progression of lower-extremity disability in 13. King, M. B. and Tinetti, M. E. (1995) Falls in older women with diabetes: the Women’s Health and community-dwelling older persons. J Am Geriatr Soc, Aging Study. Diabetes Care, 26, 70–5. 43, 1146–54. 25. Gill, T., Taylor, A. W. and Pengelly, A. (2005) A 14. Volpato, S., Leveille, S. G., Blaum, C., Fried, L. P. and population-based survey of factors relating to the Guralnik, J. M. (2005) Risk factors for falls in older prevalence of falls in older people. Gerontology, 51, disabled women with diabetes: the women’s health 340–5. and aging study. J Gerontol A Biol Sci Med Sci, 60, 26. Leipzig, R. M., Cumming, R. G. and Tinetti, M. E. 1539–45. (1999a) Drugs and falls in older people: a systematic 15. Nelson, J. M., Dufraux, K. and Cook, P. F. (2007) The review and meta-analysis: I. Psychotropic drugs. JAm relationship between glycemic control and falls in older Geriatr Soc, 47, 30–9. adults. J Am Geriatr Soc, 55, 2041–4. 27. Leipzig, R. M., Cumming, R. G. and Tinetti, M. E. 16. Adler, A. I., Stratton, I. M., Neil, H. A., Yudkin, J. S., (1999b) Drugs and falls in older people: a systematic Matthews, D. R., Cull, C. A., Wright, A. D., Turner, R. review and meta-analysis: II. Cardiac and analgesic C. and Holman, R. R. (2000) Association of systolic drugs. J Am Geriatr Soc, 47, 40–50. blood pressure with macrovascular and microvascular 28. UK Prospective Diabetes Study (UKPDS) Group complications of type 2 diabetes (UKPDS 36): prospec- (1998) Intensive blood-glucose control with sulpho- tive observational study. Br Med J, 321, 412–19. nylureas or insulin compared with conventional 17. Coleman, A.L., Cummings, S.R., Yu, F., Kodje- treatment and risk of complications in patients with bacheva, G., Ensrud, K. E., Gutierrez, P., Stone, K. type 2 diabetes (UKPDS 33). Lancet, 352, 837–53. L., Cauley, J. A., Pedula, K. L., Hochberg, M.C. and 29. Campbell, A. J., Borrie, M. J. and Spears, G. F. (1989) Mangione, C.M. (2007) Binocular visual-ﬁeld loss Risk factors for falls in a community-based prospective increases the risk of future falls in older white women. study of people 70 years and older. J Gerontol, 44, J Am Geriatr Soc, 55, 357–64. M112–17. 18. Freeman, E. E., Munoz, B., Rubin, G. and West, S. 30. Cumming, R.G. (1998) Epidemiology of K. (2007) Visual ﬁeld loss increases the risk of falls medication-related falls and fractures in the in older adults: the Salisbury eye evaluation. Invest elderly. Drugs Aging, 12, 43–53. Ophthalmol Vis Sci, 48, 4445–50. 31. Schwartz, A. V., Hillier, T. A., Sellmeyer, D. E., 19. Stratton, I. M., Adler, A. I., Neil, H. A., Matthews, D. Resnick, H. E., Gregg, E., Ensrud, K. E., Schreiner, P. R., Manley, S. E., Cull, C. A., Hadden, D., Turner, R. J., Margolis, K. L., Cauley, J. A., Nevitt, M. C., Black, C. and Holman, R. R. (2000) Association of glycaemia D. M. and Cummings, S. R. (2002) Older women with with macrovascular and microvascular complications diabetes have a higher risk of falls: a prospective study. of type 2 diabetes (UKPDS 35): prospective observa- Diabetes Care, 25, 1749–54. tional study. Be Med J, 321, 405–12. 32. Schwartz, A. V., Sellmeyer, D. E., Ensrud, K. E., 20. Vaughan, N., James, K., McDermott, D., Griest, S. and Cauley, J. A., Tabor, H. K., Schreiner, P. J., Jamal, Fausti, S. (2006) A 5-year prospective study of diabetes S. A., Black, D. M. and Cummings, S. R. (2001) and hearing loss in a veteran population. Otol Neurotol, Older women with diabetes have an increased risk of 27, 37–43. fracture: a prospective study. J Clin Endocrinol Metab, 21. Menz, H. B., Lord, S. R., St George, R. and Fitzpatrick, 86, 32–8. R. C. (2004) Walking stability and sensorimotor func- 33. Bonds, D. E., Larson, J. C., Schwartz, A. V., Strot- tion in older people with diabetic peripheral neuropa- meyer, E. S., Robbins, J., Rodriguez, B. L., Johnson, thy. Arch Phys Med Rehabil, 85, 245–52. K. C. and Margolis, K. L. (2006) Risk of fracture in 29.8 PATIENT SAFETY AND QUALITY OF CARE 415

women with type 2 diabetes: the Women’s Health Ini- and Wong, J. B. (2004) Effect of Vitamin D on falls: a tiative Observational Study. J Clin Endocrinol Metab, meta-analysis. JAMA, 291, 1999–2006. 91, 3404–10. 42. Bischoff-Ferrari, H. A., Orav, E. J. and Dawson- 34. American Geriatrics Society, British Geriatrics Society Hughes, B. (2006) Effect of cholecalciferol plus cal- and Prevention, American Academy Of Orthopaedic cium on falling in ambulatory older men and women: Surgeons Panel On Falls Prevention. (2001) Guideline a 3-year randomized controlled trial. Arch Intern Med, for the prevention of falls in older persons. JAm 166, 424–30. Geriatr Soc, 49, 664–72. 43. Broe, K. E., Chen, T. C., Weinberg, J., Bischoff-Ferrari, 35. Gillespie, L. D., Gillespie, W. J., Robertson, M. C., H. A., Holick, M. F. and Kiel, D. P. (2007) A higher Lamb, S. E., Cumming, R. G. and Rowe, B. H. (2003) dose of vitamin d reduces the risk of falls in nursing Interventions for preventing falls in elderly people. home residents: a randomized, multiple-dose study. J Cochrane Database Syst Rev, CD000340. Am Geriatr Soc, 55, 234–9. 36. Tinetti, M.E. (2003) Clinical practice. Preventing falls 44. Kiel, D. P., Magaziner, J., Zimmerman, S., Ball, L., in elderly persons. N Engl J Med, 348, 42–9. Barton, B. A., Brown, K. M., Stone, J. P., Dewkett, 37. Cumming, R. G., Thomas, M., Szonyi, G., Salkeld, D. and Birge, S. J. (2007) Efﬁcacy of a hip protector G., O’Neill, E., Westbury, C. and Frampton, G. (1999) to prevent hip fracture in nursing home residents: the Home visits by an occupational therapist for assessment HIP PRO randomized controlled trial. JAMA, 298, and modiﬁcation of environmental hazards: a random- 413–22. ized trial of falls prevention. J Am Geriatr Soc, 47, 45. Saliba, D., Elliott, M., Rubenstein, L. Z., Solomon, D. 1397–402. H., Young, R. T., Kamberg, C. J., Roth, C., Maclean, 38. Mathias, S., Nayak, U. S. and Isaacs, B. (1986) Balance C. H., Shekelle, P. G., Sloss, E. M. and Wenger, N. in elderly patients: the ‘get-up and go’ test. Arch Phys S. (2001) The Vulnerable Elders Survey: a tool for Med Rehabil, 67, 387–9. identifying vulnerable older people in the community. 39. Podsiadlo, D. and Richardson, S. (1991) The timed ‘Up J Am Geriatr Soc, 49, 1691–9. and Go’: a test of basic functional mobility for frail 46. Rubenstein, L. Z., Powers, C. M. and MacLean, C. elderly persons. J Am Geriatr Soc, 39, 142–8. H. (2001) Quality indicators for the management and 40. Boyd, C. M., Darer, J., Boult, C., Fried, L. P., Boult, prevention of falls and mobility problems in vulnerable L. and Wu, A. W. (2005) Clinical practice guidelines elders. Ann Intern Med, 135, 686–93. and quality of care for older patients with multiple co- 47. Shekelle, P. and Vijan, S. (2007) Quality indicators for morbid diseases: implications for pay for performance. the care of diabetes mellitus in vulnerable elders. JAm JAMA, 294, 716–24. Geriatr Soc, 55 (Suppl. 2), S312–17. 41. Bischoff-Ferrari, H. A., Dawson-Hughes, B., Willett, W. C., Staehelin, H. B., Bazemore, M. G., Zee, R. Y. SECTION VI Optimizing Diabetes Care in Older People

Diabetes in Old Age Third Edition Edited by Alan J. Sinclair © 2009 John Wiley & Sons Ltd. ISBN: 978-0-470-06562-4 30 The Role of Specialist Nurses and Other Members of the Multidisciplinary Team (MDT)

Carolin D. Taylor and Timothy J. Hendra Robert Hadﬁeld Wing, Northern General Hospital, Department of Geriatric Medicine, Shefﬁeld, UK

team (MDT) that has to employ additional skills to Key messages those used when caring for younger adults with less comorbidity. The aims of the MDT in caring for an • Multidisciplinary assessment of the older person older person with diabetes embrace those for a younger with diabetes should include formal evaluation adult but include, in addition, the need to take par- of cognition, mood and functional status. ticular account of these comorbidities in the context • Liaison between the multidisciplinary team of normal ageing and the need for informal and for- and the formal/informal carers is important in mal carer support (Table 30.1). As well as education order to review treatment goals, the risk of and evaluating the factors that will influence diabetes hypoglycaemia, and the need for rehabilitation in particular as the older person with diabetes management, the team may need to screen the person ages and becomes frail. for mood, cognitive dysfunction and functional status, • The roles and scope of practice of Specialist which could in turn influence the patient’s treatment Nurses and other members of the multidisci- concordance and ability to cope with hypoglycaemia. plinary team in the UK is evolving to include The person’s present and future reliance on formal and prescribing and other duties previously under- informal carers will affect the risk/benefit assessment taken by medical staff. of different therapy options, as well as treatment goals and the strategy for review. When managing younger adults, the diabetes MDT 30.1 Introduction often makes referrals to other agencies such as vascular surgeons and ophthalmologists. However, for the Older people with diabetes have a high prevalence older person with diabetes the list of potential comor- of recognized and unrecognized cognitive dysfunc- bidities and interactions with other specialist MDTs is tion, depression and functional disability [1]. This greater, with the additional need to recognize when presents particular challenges for the multidisciplinary rehabilitation may be needed.

Diabetes in Old Age Third Edition Edited by Alan J. Sinclair © 2009 John Wiley & Sons Ltd. ISBN: 978-0-470-06562-4 420 CH 30 THE ROLE OF SPECIALIST NURSES AND OTHER MEMBERS OF THE MULTIDISCIPLINARY TEAM (MDT)

Table 30.1 The aims of multidisciplinary care for severe hypoglycaemia. In this study, ambulance crews older people. who have been trained by Diabetes Specialist Nurses (DSNs) prospectively referred people who had been 1 Development of a personalized care plan for the management of the person’s diabetes that takes treated successfully at home for an acute hypogly- account of their other comorbidities. caemic episode and who were safe not to be admit- 2 Screening, prevention and treatment of acute and ted, to the local DSN team. All people referred were chronic diabetes complications. contacted by telephone, and the majority (79%) also 3 Rapid identification and referral of the person with attended a clinic appointment where their symptoms diabetes to other agencies for the management of and treatment were reviewed. The people in this study non-diabetes related problems, including were aged 25 to 92 years, and 50% of them were aged rehabilitation when necessary. >60 years [6]. 4 Maintenance of well-being of the person with diabetes with their continued residence at home in the community wherever possible. 30.3 The changing role of the MDT 5 Effective working with other specialist teams caring for older people. in service delivery 6 Recognition of the need to change the management of a person’s diabetes with normal ageing. The role of the MDT has been changing rapidly over 7 Education and effective support for the formal and recent years as professional boundaries have become informal carers of older people with diabetes. less distinct and care has become firmly person-based, with service changes aimed at reducing constraints associated with the primary/secondary care interface. 30.2 The relationship of the Within the team the roles of the DSN – and to a lesser diabetes MDT with other extent of the pharmacist – may evolve further in the UK as a consequence of their ability to act as indepen- services for older people dent and supplementary prescribers, respectively. Many changes in service delivery are, however, The diabetes MDT will need to establish effective links driven in the main to achieve biomedical targets of with other teams who may care for the same person glycaemic, blood pressure and lipid control, as well (Figure 30.1). These people may be dependent upon preventive screening of the eyes and feet. In pursuit formal and informal carers for support in order to live of these goals, people are submitted to ever-increasing independently in the community. As they approach the polypharmacy that invariably leads to problems with end of life they often become frail and unsteady, with concordance and the potential for treatment-limiting an increased risk of falling that may be exacerbated side effects such as hypoglycaemia. Some of the un- by hypoglycaemia or poor diabetes control. Falls in measured benefits of multidisciplinary care may be the older person with diabetes are also associated with lost – those of spending time with the person to evalu- reduced muscle strength and greater risk of fracture ate their knowledge deficiencies, providing information [2–5]. Due to mood and cognitive problems, elderly and encouraging health-related behaviours that may re- people with diabetes in the UK may present to Old sult in the achievement of biomedical targets, without Age Psychiatry Team in the community as well as hos- the need for a quickly written prescription. General pital, and also be supported by community psychiatric practitioners in the UK recognize that their efforts of- nurses. ten do not meet their patients’ expectations and that Other teams that may be caring for the same person the MDT may provide the tailored approach to care, may include palliative care, heart failure and cardiac encompassing their need to meet targets and improve rehabilitation, as well community matrons and interme- patient adherence/compliance with their advice [7]. diate care teams. It is important that the diabetes MDT can link effectively with other teams and know when to step forward and coordinate care, or take a back 30.4 The ‘Patient’ seat, when diabetes is not the person’s main problem. A recent study from the UK has also demonstrated Patient empowerment and role in deciding their own the value of inter-agency collaboration between the di- management and treatment goals are well established abetes service and ambulance services for people with as the centre of MDT functioning. The majority of 30.4 THE ‘PATIENT’ 421

Physiotherapist Occupational therapist

Social worker Diabetes Specialist Speech and Nurse language therapist Dietitian General Practitioner and Practice Nurse

Pharmacist Falls PERSON Podiatrist Ambulance service Service Relatives

Community Diabetologist Nurses and and/or Matrons Geriatrician Formal and Specialist informal carers Palliative stroke team care team

Heart failure Memory service Old-age clinic psychiatry team

Figure 30.1 The relationships of the diabetes multidisciplinary team (MDT). younger adults are cognitively intact and have capacity capacity, as well as the position of Advance Directives to participate in these decisions. However, many older in guiding care for someone who lacks capacity. people have a limited capacity for decision making For those people who lack or who have limited ca- as a result of dementia, mood abnormalities or cere- pacity, the MDT should liaise with relatives, friends brovascular disease. The effect of these diseases on the and perhaps formal and informal carers to identify ability of the person to make a decision may be exacer- risks of hypoglycaemia due to over-medication or in- bated by hypo- and hyperglycaemia, and may fluctuate consistent food intake, as well as the person’s ability to in severity. These issues are the centre of the Mental recognize and cope with this problem. The information Capacity Act in the UK, which clarifies the roles of rel- supplied should influence the choice of treatment, gly- atives and friends in supporting older people who lack caemic targets and any monitoring in the community. 422 CH 30 THE ROLE OF SPECIALIST NURSES AND OTHER MEMBERS OF THE MULTIDISCIPLINARY TEAM (MDT) 30.5 The roles of the individual relatives and friends, and formal carers often took sec- team members ond place to the immediate demands of the service, but is no less important. To these roles may now be added 30.5.1 The diabetes specialist nurse or that of delivering nurse-led diabetes clinics as a con- diabetes educator sequence of their ability to prescribe, thereby allowing the additional scope to manage aspects of patient care The diabetes specialist nurse (DSN) in the past has had that hitherto were the sole prerogative of medical staff. two main roles (see Table 30.2). Principally, in terms of Hospital-based DSNs have traditionally concen- the time involved, DSNs delivered or supported care by trated upon giving lifestyle advice, support and giving advice either face to face or over the telephone medicines management, either face to face or over the to people, as well as to other health care professionals. telephone. There is evidence that their role improves Previously, the role of the DSN in educating people, glycaemic control and reduces hospital length of stay

Table 30.2 Key roles for the multidisciplinary team (MDT). Team member Role 1 Diabetes Specialist Nurse (DSN) or Diabetes • Education for patients and carers Educator (Hospital or community nurse • Advice on diabetes medicines management with additional qualiﬁcations in diabetes) • Delivering specialist diabetes clinics • Diabetes related prescribing to protocol/algorithm • Education and advice for health care professionals

2 Nurses other than DSNs • Management of non-diabetes problems • Non-diabetes prescribing to protocol/algorithm

3 Dietician • Assessment of nutritional status • Advice on healthy eating • Advice to older people with malnutrition • Diabetes-related dietary advice

4 Podiatrist • Education for self-examination of feet and appropriate footwear • Advice for age-related foot problems • Screening, prevention and treatment of diabetes-related foot problems

5 Community Pharmacist • Educational support • Recognition of iatrogenic problems • Medication review clinics (UK) • Diabetes related prescribing to protocol/algorithm (USA and UK)

6Doctor • Advice on diabetes medicines management • Delivering specialist diabetes clinics • Service development and clinical governance • Audit 30.5 THE ROLES OF THE INDIVIDUAL TEAM MEMBERS 423 [8–10]. DSNs employed specifically to manage older more recently specifically around carbohydrate count- adults with diabetes, either in the acute setting or ing. In the older person it may be much more important community, play a vital role in meeting the changing to assess overall nutritional status, especially follow- needs of this increasing population. ing, for example, a cerebrovascular accident when the Recently, changes in the United Kingdom have al- person’s ability to eat may be severely affected. The lowed DSNs to become independent prescribers; pro- dietician can advise on a balanced diet and/or recom- vided that the medicines are within their area of mend dietary supplements to ensure adequate nutri- competence, independent prescribers may prescribe tional intake to both stabilize glycaemia and prevent any licensed medicine and some controlled drugs. potential complications of malnutrition, such as pres- Hospital-based DSN prescribers have been shown to sure ulcers (see Table 30.2). reduce medication errors and to have a beneficial effect on length of stay [11]. 30.5.4 The podiatrist In the USA, Diabetes Managed Care Programs can Good footcare is important for all older people, be delivered by specially trained nurses and pharma- whether or not they have diabetes. Ageing is associated cists, using detailed protocols and algorithms under the with a thinning of the skin and loss of elasticity and supervision of a diabetologist. In both middle-class well as degenerative joint problems. With appropriate populations and populations of mixed ethnic back- advice from a chiropodist/podiatrist, growing old need grounds, the American Diabetes Association process not be associated with discomfort, pain and poor and outcome measures are more likely to be met in mobility due to some common foot problems. Podiatry people receiving nurse-directed care compared to those services also play a significant role in improving the receiving usual care from busy physicians [12–14]. quality of life and reducing the long-term complica- Whereas, physicians are not good at completing guide- tions associated with diabetes. As a consequence, their lines due to time constraints, trained nurses are good role in preventing diabetic amputations can save the at sticking to protocols and have more time to provide UK National Health Service an average of £80 000 lifestyle advice and preventive care. In the USA there per person. Podiatrists have a role in the treatment is a clear trend for preventive care to be delivered by of hard skin and callus, as well providing nail care. non-physician clinicians [15], and this has been rein- They can teach older people how to care for and forced by the results of a recent study which showed examine their feet regularly, to look for changes in that nurse-directed diabetes care in a US county public colour, swelling, skin damage, redness and discharge. health clinic reduced urgent hospital visits and admis- In addition, they can provide advice on footwear. sions for preventable diabetes-related causes compared to usual care [16]. 30.5.5 The community pharmacist 30.5.2 Nurses other than DSNs Elderly people often approach community pharmacists for help and advice with non-diabetes-related prob- The ability of nurses to prescribe according to protocol lems. These interactions provide the opportunity for has led to the development of nurse-led clinics in many diabetes-related advice about lifestyle, the use of ex- specialties in the UK. For people with diabetes at high isting medications, and the need to consult other mem- risk of vascular disease, nurse-led clinics employing bers of the MDT, as well for the purchase of new a systematic approach to blood pressure control with over-the-counter medications or devices. In addition, a a pre-set drug protocol can provide adequate surveil- community-based pharmacist may be the first to rec- lance and achieve blood pressure targets with reduction ognize new adverse reactions to medications and, in in cardiovascular risk [17, 18]. It is possible that the particular, any previously unrecognized hypoglyaemia. beneficial effect was related to both improved compli- In the UK, pharmacists may also participate in primary ance as well to a greater number of treatment changes; care medication review clinics. however, the challenge is to translate these studies into In the USA pharmacists can contribute to interdis- consistent practice in the long term. ciplinary health management programmes where their interventions include education and comprehensive 30.5.3 The dietician medication management through collaborative practice Within the MDT, dieticians have a traditional role of agreements with physicians. This allows them to ini- providing healthy eating guidelines or advice on diet, tiate, adjust or discontinue pharmacotherapy, to order 424 CH 30 THE ROLE OF SPECIALIST NURSES AND OTHER MEMBERS OF THE MULTIDISCIPLINARY TEAM (MDT) investigations and to make podiatry referrals. In this withdraw while remaining a point of reference for situation, pharmacists have been shown to be effective unexpected difficulties. in meeting biomedical targets of glycated haemoglobin levels and lipids, as well as improving adherence to preventive care measures such as annual eye and foot 30.6 The role of the MDT in specific inspections [19]. situations

30.5.6 The doctor 30.6.1 Age-related frailty Traditionally, the role of the doctor was often to dis- Frail elderly people sometimes have relatively small cuss management plans with the MDT at a weekly meals or eat at times that are different to when they meeting at a fixed time and place. However, in the UK were younger. This may be due to the increased time the duties and responsibilities of some diabetologists, it takes to prepare food or to a lack of nutritional geriatricians and GPs are changing significantly, such drive, but is frequently due to unrecognized depres- that it is usually impractical and unnecessary for them sion or cognitive impairment. It is important for the to be personally involved in discussing the manage- diabetes MDT to recognize when a previously fit older ment of the majority of people cared for by the MDT. person with diabetes is becoming frail, as this may The doctor’s clinical role as part of the MDT often be unknown to other agencies and it may become the cannot be restricted to a fixed time with the need to be role of the diabetes team to assess the patient’s mood, available to discuss problems with other MDT mem- cognition and functional ability, or to make a referral bers as issues arise and decisions need to be made on a to another agency to perform these assessments. Fre- daily basis. A doctor’s roles of leadership and clinical quently the onset of frailty is insidious, and with it responsibility today often require a flexible pattern of comes a significantly increased risk of hypoglycaemia, working that allows him/her to respond quickly and to both in terms of the incidence of low blood sugar support other team members as the need arises. levels and an inability of the person to manage the The doctors’ role frequently involves being responsi- problem – if he or she recognizes the symptoms. The ble for governance of the diabetes service that incorpo- occurrence of falls should indicate a need to urgently rates the implementation of clinical protocols, auditing review the person’s glycaemic control and targets, as the processes of care and the achievement of clinical well as to make the appropriate referral to a falls team and managerial outcomes by the MDT. In the past, the or for rehabilitation. doctor would sometimes personally receive most of the The onset of frailty – with or without altered mood referrals of people with diabetes to the MDT, but today or cognitive function – frequently necessitates the need most MDTs in the UK have a single point of contact for additional support at home from either family or that bypasses the doctor and allows other team mem- formal/informal carers, or perhaps the relocation of a bers to deal with issues quickly, rather than waiting for person to sheltered housing or a care home. With the the person’s details to be passed on. change in circumstances it is useful for the relatives, Many GPs in the UK have little experience, and carers and or warden to have a contact number for the are understandably wary, of initiating insulin in the diabetes team at times of crisis, or if the person is ‘off community for people with type 2 diabetes. Despite colour’ and there is concern that this could be related this there is increasing pressure from commissioners, to diabetes per se, or due to another problem such as as well as expectations from people with diabetes, for a chest/urinary infection that is being treated by the this treatment to be started outside of hospital. Many primary care physician and may affect the person’s elderly people with diabetes will be unable to attend food intake or diabetes control. a hospital clinic or GP surgery, and will expect the Sometimes, the presence of diabetes with the need service to come to them. The DSN has a unique role to manage increasing hyperglycaemia, or to avoid in this situation in terms of being able to support a recurrent hypoglycaemia, may be seen as a ‘tipping district or GP surgery nurse in educating the patient point’ for a previously independent frail person with and supervising treatment in the early stages. Once diabetes to be admitted to a care home. In this situation the patient is stabilized on insulin, and the community the DSN may be able to offer alternative options for services have developed confidence with monitoring managing the person’s diabetes, or provide additional and reviewing glycaemic targets, then the DSN can support for formal and informal carers as well as their 30.7 SUMMARY 425 primary care physician, sufficient for a package of care relevance to older people with diabetes, many of whom in the community to be sustained. As a rule, the issues may neither speak nor understand English. This prob- associated with managing diabetes should never be the lem has been studied specifically in younger adults, but reason for admission to a care home per se. not in older people. In the United Kingdom Asian Study Group an 30.6.2 Care home residents intervention of enhanced care using Asian link workers and extra community DSN sessions was In the UK there are widespread concerns about the associated with reductions in blood pressure and total quality of diabetes care in residential and nursing cholesterol, but there was no significant effect on homes that encompass the training needs of the for- glycaemic control [23]. mal carers and lack of structured review. Many care home residents are frail, with physical and cognitive impairments that make them vulnerable to poor nutri- 30.7 Summary tion and neglect [20]. Weight loss and protein-calorie under-nutrition are common, and will adversely affect DSNs and the diabetes MDT need to work effectively diabetes care. Very often, diabetes and its complica- with other specialist teams caring for older people tions are not the only chronic disease process present, with multiple pathology and comorbidities as well as yet the optimal treatment of blood glucose levels may diabetes. significantly improve a resident’s quality of life. The The roles of the DSN and MDT differ when caring diabetes MDT can act as a point of reference for the for the older person with diabetes compared to the referral of diabetes-related problems, as well as ensur- younger adult in relation to the need to routinely assess ing that the formal carers have access to education and cognitive function, mood and functional ability when support. setting treatment goals and assessing hypoglycaemic risk. 30.6.3 Terminal illness Although DSNs have always had a major role in education and prevention, in the UK the devel- People with diabetes approaching the end of their lives opment of extended practice – and prescribing in due to cancer, cardiac failure or other incurable disease, particular – means that they may take on more tasks frequently present a challenge to health care profes- previously performed by medical staff, including sionals. Often, the management of diabetes in this sit- the routine management of glycaemic control and uation is haphazard and left to clinical staff who do not cardiovascular risk factors, including hypertension have any particular expertise in diabetes management. and dyslipidaemia. A study from Cheltenham in the UK found that there The basis for caring for the older person with dia- was variability in diabetes monitoring, and little record betes remains thorough assessment, appropriate treat- of discussion with the person or family in this situa- ment goals, and structured care that accommodates tion. A significant number of people continued to have changes in the person’s cognitive and functional abil- their blood glucose monitored up until the day they ity, while referring them for rehabilitation when nec- died [21]. In another study conducted in Australia, the essary. results emphasized the lack of evidence-based guidelines for managing diabetes in the context of palliative care, and recommended their development as well as References the importance of increased collaboration between the diabetes MDT and palliative care specialists [22]. 1. Munshi M, Grande L, Hayes M, Ayres D, Suhl E, Capelson R, Lin S, Milberg W and Weinger K. Cogni- 30.6.4 People from wide ethnic tive dysfunction is associated with poor diabetes con- backgrounds trol in older adults. Diabetes Care 2006; 29: 1794–9. 2. Park SW, Goodpaster BH, Strotmeyer ES, Kuller LH, There is concern that people from non-Caucasian eth- Broudeau R, Kammerer C, de Rekeneire N, Harris TB, nic backgrounds have specific needs that pose partic- Schwatz AV, Tylavsky FA, Cho Y-W and Newman AB. ular challenges to the diabetes MDT. The issues of Accelerated loss of skeletal muscle strength in older greater deprivation and social isolation, as well as di- adults with type 2 diabetes; the health aging and body etary needs and cultural beliefs, may be of particular composition study. Diabetes Care 2007; 30: 1507–12. 426 CH 30 THE ROLE OF SPECIALIST NURSES AND OTHER MEMBERS OF THE MULTIDISCIPLINARY TEAM (MDT)

3. Schwartz AV, Hillier TA, Sellmeyer DE, Resnick HE, 13. Taylor CB, Miller NH, Reilly KR, Greenwald G, Cun- Gregg E, Ensrud KE, Schreiner PJ, Margolis KL, ning D, Deeter A and Abascal L. Evaluation of a nurse Cauley JA, Nevitt MC, Black DM and Cummings SR. care management system to improve outcomes in peo- Older women with diabetes have a higher risk of falls: ple with complicated diabetes. Diabetes Care 2007; 26: a prospective study. Diabetes Care 2002; 25: 1749–54. 1058–63. 4. Bonds DE, Larson JC, Schwartz AV, Strotmeyer ES, 14. Davidson MB. Effect of nurse-directed diabetes care Robbins J, Rodriguez BL, Johnson KC and Margolis in a minority population. Diabetes Care 2003, 26, KL. Risk of fracture in women with type 2 diabetes: 2281–7. the Women’s Health Initiative Observational Study. J 15. Druss BG, Marcus SC, Olfson M, Tanielian T and Clin Endocrinol Metab 2006; 91: 3404–10. Pincus HA. Trends in care by non-physician clinicians 5. Lipscombe LL, Jamal SA, Booth GL and Hawker in the United States, N Engl J Med 2003; 348: 130–7. GA. The risk of hip fractures in older individuals 16. Davidson MB, Ansari A and Karlan VJ. Effect of a with diabetes: a population-based study. Diabetes Care nurse-directed diabetes management program on urgent 2007: 30: 835–41. care/emergency room visits and hospitalisations in a 6. Walker A, James C, Bannister M and Jobes E. Evalua- minority population. Diabetes Care 2007; 30: 224–7. tion of a diabetes referral pathway for the management 17. Singh PK, Beach P, Iqbal N, Buch HN and Singh BM. of hypoglycaemia following emergency contact with Nurse-led management of uncontrolled hypertension in the ambulance service to a diabetes specialist nurse those with diabetes and high vascular risk. Pract Diab team. Emerg Med J 2006; 23: 449–51. Int 2007; 24: 92–6. 7. Wens J, Vermeire E, Royen PV, Sabbe B and Denekens 18. Denver EA, Barnard M, Woolfson RG and Earle KA. J. GPs perspectives of type 2 diabetes people’ adher- Management of uncontrolled hypertension in nurse-led ence to treatment: A qualitative analysis of barriers and clinic compared with conventional care for people with solutions. BMC Fam Pract 2005; 6: 20 type 2 diabetes. Diabetes Care 2003; 26: 2256–60. 8. Koproski J, Pretto Z and Poretsky L. Effects of an 19. Brooks AD, Rihani RS and Derus CL. Pharmacist intervention by a diabetes team in hospitalised people membership in a medical group’s diabetes health man- with diabetes. Diabetes Care 1997; 20: 1553–55. agement program. Am J Health-Syst Pharm 2007; 64: 9. Thompson DM, Kozak SE and Sheps S. Insulin ad- 617–21. justment by a nurse educator improves glucose control 20. Morley JE and Silver AJ, Nutritional issues in nursing in insulin-requiring diabetic people: a randomised trial. home care. Ann Intern Med 1995; 123: 850–9. Can Med Assoc J 1999; 161: 959–62. 21. McCoubrie R, Jeffrey D, Paton C and Dawes L. Man- 10. Wong FKY, Mok MPH, Chan T and Tsang MW. aging diabetes mellitus in people with advanced cancer: Nurse follow up of people with diabetes: randomized a case note audit and guidelines. Eur J Cancer Care controlled trial. J Adv Nur 2005; 50: 391–402. 2005; 14: 244–8. 11. Courtenay M, Carey N, James J, Hills M and Roland 22. Quinn K, Hudson P and Dunning T. Diabetes man- J. An evaluation of a specialist nurse prescriber on di- agement in people receiving palliative care. J Pain abetes in-person service delivery. Pract Diab Int 2007; Symptom Manage 2006; 32: 275–86. 24: 69–74. 23. O’Hare JP, Raymond NT, Mughal S, Dodd L, Hanif W, 12. Aubert RE, Herman WH, Waters J, Moore W, Sut- Ahmad Y, Mishra K, Jones A, Kumar S, Szczepura A, tonD,PetersonBL,BaileyCandKoplanJP.Nurse Hillhouse EW and Barnett AH. UKADS Study Group. case management to improve glycaemic control in di- Evaluation of delivery of enhanced diabetes care to abetic people in a health maintenance organization: a people of South Asian ethnicity: the United Kingdom randomized trial. Ann Intern Med 1998; 129: 605–12. Asian Diabetes Study. Diabet Med 2004; 21: 1357–65. 31 Diabetes Education in the Elderly

Charles Fox and Anne Kilvert Diabetes Centre, Northampton General Hospital, Northampton, UK

involve RCTs with designs based on explicit Key messages hypotheses and educational theory, and include a range of outcomes evaluated after long follow-up • As with younger people, diabetes education is intervals. Studies should aim to determine the an essential intervention to enhance care in the characteristics, in terms of type, length and frequency elderly. of intervention, team composition and setting, that • In its simplest form, diabetes education is a would maximise the impact of patient education in learning process that requires a positive attitude both the short and longer term. Such studies should in older people, with regular reinforcement and also include qualitative evaluation of the educational engagement. intervention itself, and research to identify the • Group education is increasingly regarded as a characteristics of education that are most important more effective method of educating people with for different stages of the disease, and that best diabetes. match different cultural and social needs.” • In order for diabetes education in the el- As a result of the renewed focus on education, a derly to be effective, several patient-related and number of courses have been developed in various professional-related barriers need to be over- formats, but DAFNE, X-PERT and DESMOND are come. among the few which have been validated. Even less has been published on the specific chal- 31.1 Introduction lenge of education for the elderly. The joint Dia- betes UK and Department of Health Report from the Although today, diabetes education is regarded as cen- Patient Education Working Group (‘Structured Educa- tral to diabetes care, very little published evidence is tion in Diabetes’ [2]) identifies several groups, where available to demonstrate its effectiveness. The National there is an education gap. This includes children, ado- Institute for Health and Clinical Excellence (NICE) lescents and their carers, pregnant women, and vulnerable groups such as those with learning disabilities provides ‘Guidance in the use of patient education and limited learning and language skills. There is no models in diabetes’ [1], which includes the following mention in this exhaustive document of the special ed- statement: ucational needs of elderly people. Much of the present “The paucity of high-quality trials of the effectiveness comment on the specific needs of the older person of patient-education models for diabetes, particularly is derived from the personal observations of people those for people with type 2 diabetes, reveals a who have delivered educational courses to people of need for more research. Further research should all ages. The main focus will be on type 2 diabetes,

Diabetes in Old Age Third Edition Edited by Alan J. Sinclair © 2009 John Wiley & Sons Ltd. ISBN: 978-0-470-06562-4 428 CH 31 DIABETES EDUCATION IN THE ELDERLY although the aspects of type 1 diabetes that are partic- fresh air and activity and mental exercise of taking ularly relevant to the elderly will also be discussed. a few pictures makes a big difference to me when I am in need of perking up. Eat fish, lean chicken and low GI carbs, and walk 10,000 steps a day. Works for 31.2 Principles of adult diabetes me.” education Older people may need some encouragement to enter the website, but they can join the army of ‘silver Successful education for adults is a process of learn- surfers’ who know their way round the internet. ing (by the adult) rather than teaching (by the expert). These principles apply as much to diabetes as any other form of learning. We quote the Joslin Clinic 31.3 How should education be principles for diabetes education as an example of delivered? person-centred diabetes care from a leading US diabetes centre with a longstanding reputation for patient 31.3.1 Evidence for effectiveness of group education. The Joslin Clinic was founded by Elliott education Joslin (1869–1962), the first doctor in the USA to spe- cialize in diabetes, and an advocate for patient educa- Traditionally, education has been delivered in a tion. The Joslin Clinic is a leader in promoting diabetes one-to-one, didactic fashion, with the agenda deter- self-management education. mined by the experienced health care professional. However, there is increasing evidence for the effective- 31.2.1 The Joslin Clinic principles ness of group education, with the agenda determined by the participants as a ‘discovery learning’ process. • The person with diabetes is the centre of her/his The strongest evidence for group education comes health care team. from two European studies conducted by Trento et al. in Italy [3] and by Kultzer et al. Germany [4]. • People with diabetes live multifaceted lives with In Turin, Italy, Marina Trento and colleagues set competing demands that influence their diabetes up a programme in 1998 to test the effectiveness and self-care. feasibility of a carefully designed long-term group edu- • People living with diabetes make complex self-care cation programme [3]. A total of 112 people with type decisions every day. 2 diabetes was randomly allocated to either group or individual education. The groups were approximately • Family and other support systems strongly influence the same age (60 years), with a mean known duration diabetes self-care. of diabetes of 9.5 years. The groups met every three • People with diabetes learn ideas and concepts that months, and each session was facilitated by an edu- they perceive as important. cationalist and a physician. The educational objectives were to: • Learning occurs when the individuals are engaged. • reach a desirable body weight • Learning is a process that requires reinforcement and flexibility. • learn to shop for food (reading labels for content, Patients and professionals interested in interactive ed- energy values, etc.) ucation should log onto the Joslin Clinic website • choose an appropriate quality and quantity of food, (www.joslin.org). This is an active, simple to access both at home and when eating out website for diabetes information with a lively discus- • sion board. It has an American flavour – that is, upbeat increase physical activity, when feasible and honest about feelings. • take medication properly and regularly “Around lunch time today I was feeling kind of icky • know the meaning of the main laboratory tests of and yucky and hi blood sugary, so I took two hours off metabolic control work and grabbed my pocket camera and hit the trail along the beaches between False Creek and Wreck • recognize early symptoms of, and be able to react Beach here in Vancouver. Now I feel a lot better. The to, hypoglycaemia 31.4 EMPOWERMENT IN DIABETES CARE 429 • take appropriate action in case of intercurrent The groups were assessed for glycaemic control, illnesses knowledge, psychological well-being and self-care behaviour at 3 and 15 months after the intervention. • care for the feet and buy appropriate footwear While knowledge was seen to improve equally in • regularly attend clinic and screening checks for com- all groups, HbA1c levels were unchanged in the plications. didactic group but fell by 0.7% at 3 months in both self-management groups. This fall was not sustained in Each session was structured into four phases: (1) wel- group C, but was maintained at 15 months in group B. come and introduction to the subject; (2) interactive Both psychological well-being and self-care behaviour learning; (3) discussion of patient experiences; and (4) were significantly higher in the self-management conclusions with directions for ‘homework’ and med- groups, but there was no added benefit from the ical consultation with the physician, if necessary. individual sessions. The authors concluded that During phase 1, patients were given sealed en- self management training had a significantly higher velopes containing the results of blood tests, but these medium-term efficacy, and that group sessions were were only discussed collectively if the patients so de- more effective than an individual approach sired. During phases 2 and 3, various hands-on activities, group work, problem-solving exercises, real-life 31.3.2 Comparison of one-to-one and simulations and role play were carried out. In order to group education induce positive group dynamics, the facilitators worked The group effect can be very powerful, particularly with each patient to identify and share problems and if it is well facilitated. This outweighs the obvious successes with other group members. This helped the drawbacks of having to coordinate group sessions group dynamics, and any examples of unhelpful be- (see Table 31.1). Group discussions allow patients and haviour were not criticised but rather used as a source their carers to share experiences and feelings, and to of positive learning for the group. The control group identify with stories that others bring to the group. received traditional one-to-one consultations and edu- Discussion with others in the same situation often cation sessions. reinforces the educational message. However, people The groups were maintained over five years and, with specific disabilities may have difficulty coping while the control group showed a steady and expected with the group setting. rise in HbA1c level over time, those patients taking The advantages and disadvantages of individual and part in group care maintained a steady HbA1c level group education are compared in Table 31.2. (Table 31.1). In Bamberg, Germany, Bernhard Kulzer and 31.4 Empowerment in diabetes care colleagues [4] compared three different treatment/ education programmes to evaluate the didactic versus It is now widely accepted that the only realistic way to self-management approach as well as the group effect: achieve good control of diabetes is to enable the person • Group A received a didactic intervention focused on with diabetes to make their own decisions. This is an teaching knowledge, skills and information about the integral part of the Department of Health philosophy, correct treatment of diabetes. This was delivered as and underpins the National Service Framework for Di- four, 90-min sessions in a group setting. abetes [5]. Standard 3 states that: “All children, young people and adults with diabetes will receive a service • Group B took part in a self-management/em- which encourages partnership in decision-making, sup- powerment programme focused on emotional, ports them in managing their diabetes and helps them cognitive and motivational aspects of behaviour to adopt and maintain a healthy lifestyle. This will be change to promote lifestyle modifications, especially reflected in an agreed and shared care plan in an appro- with regards to eating and exercise. This consisted priate format and language. Where appropriate, parents of 12, 90-min sessions in a group setting. and carers should be fully engaged in this process.” • Group C was also a self-management/empowerment The NSF Standards document goes on to specify how empowerment should be promoted: programme with the same curriculum as group B, but delivered as 12 sessions – six individual and six • The provision of information, education and psy- in a group. chological support that facilitates self-management 430 CH 31 DIABETES EDUCATION IN THE ELDERLY

Table 31.1 Biochemical and clinical variables at baseline and year 5 in patients followed by group care (n = 42) and control subjects (n = 42). Baseline Year 5 Change from P-valuea baseline Knowledge of diabetesb Group 15.5 27.9 12.4 <0.001 Control 21.4 18.0 −3.4 Problem-solving abilityb Group 11.4 17.0 5.6 <0.001 Control 12.3 10.0 −2.3 Quality of lifeb Group 67.4 43.7 −23.7 <0.001 Control 70 89.2 19.2 HbA1c (%) Group 7.4 7.3 0 <0.001 Control 7.4 9.0 1.6

a Difference between group care and control subjects. bDetails of the measures used are provided in Ref. [3]. Table 31.2 Comparison of one-to-one and group • A family and social environment that supports be- education. haviour change: families and communities provide both practical support and a framework for the in- One to one Group dividual’s beliefs. Timing Easy to organize Requires • The tools to support behaviour, for example, af- coordination fordable healthier food options both at home and Location No special Limitedbyspace in the workplace. requirements and accessibility • Format Tailored to Determined by Active involvement in negotiating, agreeing and individual group – different owning goals. perspectives • Knowledge to understand the consequences of Adaptability Easy to adjust for May be unsuitable different choices and to enable action. patient disability for those with physical or Self-management can only be successful if the per- mental disability son feels empowered to make their own decisions. The Agenda Individualized and Enriched by group empowerment model was developed in the US by An- may be limited derson and Funnell [6], and is one of many approaches Educator training Adult education Group education designed to encourage the individual to work out their skills skills own management plan and implement their own deci- Added benefit Intimacy, Group experience sions. Traditional diabetes care is often disempowering, personalized reinforces as it instructs people on the changes they need to make learning Efficiency Limited by More time in order to achieve good control, without giving them available time available for the opportunity to identify the changes they would learning choose for themselves. People may feel pressurized to Patient’s Security/safety Social anxiety of agree to changes to keep the health professional happy, anticipation group exposure but if they do not perceive a personal benefit, they will not generate the motivation to change. is therefore the cornerstone of diabetes care. People The structure of UK health care is weighted in with diabetes need the knowledge, skills and motiva- favour of the professionals. Clinics are arranged around tion to assess their risks, to understand what they will the convenience of doctors, and the people are rou- gain from changing their behaviour or lifestyle and to tinely kept waiting. In secondary care, the consul- act on that understanding by engaging in appropriate tation may be with an unfamiliar health care pro- behaviours. Other beneficial factors include: fessional, who sees their role as an advice-giver. In primary care, the hard-pressed practice nurse may 31.4 EMPOWERMENT IN DIABETES CARE 431 feel obliged to focus on NHS targets, rather than the of disease and pathology, but suggests that these should expectations and needs of the individual patient. The always be considered in parallel with – and in equal following items are examples of common practice, importance to – the patient’s individual experiences of though many health care professionals try very hard their condition (see Figure 31.1). This specifically in- to provide patient-centred care: cludes eliciting the individual’s ideas, concerns and expectations of their condition and treatments. • Primary Care ◦ Focus on annual review and targets ◦ Computer screen demanding attention of health 31.4.3 Care planning model care professional A Diabetes Care Planning Working Group was set ◦ Template rather than patient-focussed up by the Department of Health and Diabetes UK to ◦ Pressure to achieve Quality and Outcomes Frame- facilitate this process in the UK. The group developed work (QOF) points a model for effective care planning which should • Secondary Care be incorporated into routine diabetes care [7]. This ◦ Large clinics and waiting rooms draws on research in clinical practice, psychology and ◦ Multiple health care professionals education to set out a process of negotiation and shared ◦ Poor continuity of care decision making between the health care professional ◦ ‘Clinical’environment and the person with diabetes. Family members or ◦ Lack of privacy carers should be involved in this process, if possible (Figure 31.2). • Access The components of the model include: ◦ Appointments centred on convenience of the or- • ganization The individual’s story and the professional’s story ◦ Travel and parking • Possible topics for discussion 31.4.1 Who should be empowered? • Learning about diabetes All people who are capable of making their own de- • Managing diabetes cisions about self-management should be encouraged • Living with diabetes to set their own goals, and to decide how these are to be achieved. The priorities may not necessarily be • Other health and social issues those of the health care professional. This approach is • Sharing and discussing information and negotiating embodied in the Care Planning process, which is being the agenda introduced by the Department of Health for all people with long term-conditions. • Action planning • Documentation 31.4.2 The principles of care planning for diabetes and other long-term Each of these topics is discussed in detail in the conditions Department of Health/Diabetes UK booklet. Care planning is underpinned by the principles of 31.4.4 Empowerment in the elderly patient-centredness and partnership working [7]. It is an ongoing process of two-way communication, ne- While there is good evidence for the effectiveness of gotiation and joint decision-making, in which both the empowerment approach in general diabetes care, the person with diabetes and the health care profes- its value in the elderly cannot be assumed. This has sional make an equal contribution to the consulta- been evaluated in a group of 148 Cuban subjects aged tion. It differs from the ‘paternalistic’ or ‘health care >60 years who participated in a 5-year study of Con- professional-centred’ model of consulting, traditionally tinuing Interactive Education (CIE), which followed on applied in acute settings. from the established basic diabetes information course The Disease-Illness model has been proposed as a [9]. Sixty interactive meetings, each lasting for approx- means of achieving this [8]. This emphasizes the im- imately 90 minutes and ranging in format from group portance of the health care professional’s perspective discussions to cultural activities, meals out and formal 432 CH 31 DIABETES EDUCATION IN THE ELDERLY

Patient presents problem

Gathering Information

Parallel search of two frameworks

illness framework Disease framework Patient's agenda: Doctor's agenda ideas symptoms concerns signs expectations investigations feelings underlying pathology thoughts effects

Understanding the patients's unique Differential diagnosis experience of illness

Integration of the two frameworks

Explanation and planning: shared understanding and decision-making

Figure 31.1 Parallel history-taking in care planning.

conferences, were held at monthly intervals over a Not all older people are able to benefit from such ed- 5-year period. The aim was to empower patients by ucational techniques, and Garcia and Suarez excluded encouraging skills and perceptions required to cope those with deafness, psychiatric disease and other con- with diabetes, rather than to provide clinical informa- ditions, such as stroke, which may interfere with their tion. Participants were invited to select the topics for ability to comprehend. Elderly people need to be as- discussion and to work in several small groups (up to sessed to identify factors which may interfere with their 15 people) to identify problems and arrive at solutions. capacity for empowerment. The health care professionals acted as facilitators, but 31.4.5 Barriers to empowerment in the not as information givers. The small groups then united elderly to exchange ideas and enrich discussion. A comparison of results pre-programme and 5 years Such barriers might derive from the older person them- later showed significant reductions in HbA1c level selves, or from the professionals working with or car- (from 12.4 to 7.9%; p<0.02), body weight and medi- ing for them. There is a tendency to treat older people cation requirements, together with an increased adher- as dependents who are not capable of making their own ence to self-care strategies (e.g. diet, exercise, foot- decisions: care) and a reduction in the number of diabetes-related • Barriers from the elderly person conditions requiring emergency services and hospital ◦ confused/forgetful admission. There was a significant increase in knowl- ◦ hearing or visual impairment edge and skills scores (p < 0.001), and the prevalence ◦ reluctance to make decisions for themselves – of depression fell from 69% to 24%. Although the “You just tell me what to do” study lacked a control group, it demonstrated clearly ◦ ingrained respect for authority – “I will do what that this method was effective in encouraging diabetes you tell me” self-management in older people. ◦ may be less assertive/afraid to ask 31.4 EMPOWERMENT IN DIABETES CARE 433

professional's Individual's story story

Learning Living Other health Managing about wtih and social diabetes diabetes diabetes issues

Share and discuss information

Negotiate agenda

Action Action Action Action

Figure 31.2 The care-planning process.

◦ may believe that nothing can be done to improve • Speaking louder and more slowly the situation • Using limited vocabulary ◦ may be happy to maintain the status quo; an informed decision not to change • Repeating or paraphrasing what has just been said ◦ may not be in a position to make decisions (e.g. • residential homes) – no control over food intake Using terms of endearment out of context (e.g. ‘honey’, ‘sweetie’, ‘dearie’, ‘pop’) • Barriers from professionals • ◦ ageism – assumption that the individual cannot Referring to a person in the plural – “Are we ready make their own decisions for our bath?” ◦ protection – “keep things simple” – “I’ll tell you • Using patronizing statements that sound like what to do” questions – “You would rather wear blue socks, ◦ Elderspeak (see below) wouldn’t you?”

31.4.6 Elderspeak Elderspeak is commonly heard in nursing homes, hos- The concept of Elderspeak was taken up in 1991 pitals and other settings where frail elders are found. It by Susan Kemper [10], Professor in Psychology and may be used indiscriminately with anyone who looks Gerontology at the University of Kansas. Professor old or dependent, and stereotypes the person at the re- Kemper published widely in the field of language and ceiving end. It may also be heard in everyday places ageing, and set out to develop speech modifications such as banks and supermarkets, where the old person to improve communication with the elderly, which is clearly able to function normally. she called ‘Elderspeak’. However, the term Elderspeak Elderspeak may be harmful as it implies that the has come to describe adjustments well-meaning people older person is not competent, and therefore any com- make when addressing someone who is old. These may munication problems are their fault. This language pat- be both helpful and unhelpful and include: tern affects an older person’s evaluation of their own abilities, reinforces negative stereotypes about aging, • Using a ‘singsong’ voice, changing pitch and tone, and also erodes their self-confidence. Thus, the use of exaggerating words Elderspeak may promote the very problem it is trying • Simplifying the length and complexity of sentences to avoid. 434 CH 31 DIABETES EDUCATION IN THE ELDERLY

Table 31.3 Meeting patient expectations in the consultation: Results from UK patients compared with European mean. Patient scores Mean Range UK participants (n = 625) (n = 81) Reported consultation length (min) 16.0 12.4–20.1 12.4 Patient enablement score (1–12) 5.5 3.9–6.1 3.9 Preference for involvement (%) 82.6 74.6–87.7 87.7 Evaluation of involvement (%) 70.6 55.6–81.6 55.6

Nevertheless, some important positive messages enablement, but the patient preference for involvement have emerged from studies of Elderspeak; namely that is variable and unpredictable, and so needs to be older people will understand you better if you: (i) use explored with the individual. short sentences; (ii) simplify – but be explicit; and (iii) It was of interest to see that the 81 patients re- repeat and paraphrase. cruited from the UK reported the shortest consulta- The topic of Elderspeak is reviewed in an excellent tion time, lowest enablement score, highest preference article available as a PDF file [11]. for involvement and lowest evaluation of involvement (Table 31.3), which suggested that the UK might lag 31.4.7 What do older people feel about behind their European counterparts in meeting the empowerment? patients needs and expectations. There is a feeling that older people are socialized to expect a doctor or nurse to give them advice, which 31.5 The development of structured they will then carry out to the letter. education: What to teach For example, Wensing et al. [12] investigated the factors associated with patient enablement – defined as people about diabetes the ability to cope with life and illness – in 625 elderly In the US, the NIH/US National Library of patients (aged ≥70 years) in seven European countries. Health programme for diabetes education deliv- Enhancing enablement is regarded as an important part ers essential information first, followed by more of health care, and relies on effective communication complex lifestyle and ‘special situations’ advice. between the patient and the clinician. However, pa- The programme provides an interactive web-based tient characteristics may influence their contribution to tutorial for people with diabetes: www.nlm.nih.gov/ decision-making in the consultation and elderly people medlineplus/tutorials/diabetesintroduction/, and is may be less assertive, thereby reducing their involve- based on three stages: ment in their own care. The study investigated the link between enablement, the patient’s evaluation of their • Basic “survival skills” involvement in the consultation, and their preference ◦ Dealing with hypoglycaemic episodes (‘hypos’) for involvement following a visit to their primary care ◦ Dealing with high blood sugar levels physician. ◦ Selecting the right food Immediately after the consultation, patients were ◦ How to take tablets or insulin for diabetes asked to complete validated questionnaires to assess ◦ How to test and record blood glucose levels enablement, evaluation of involvement and preference • Home management for involvement. Enablement was scored on a scale ◦ How to adjust insulin/food before exercise of1to12(high= more enablement). The mean ◦ How to handle ‘sick’ days age of the participants was 77 years and 63% had a ◦ Foot care chronic disease. The results showed the enablement ◦ How to watch for complications and associated score to be higher in those patients who reported a conditions, e.g. blood pressure, cholesterol more positive evaluation of their involvement, and this was increased if they expressed a high preference • Lifestyle changes for involvement. Thus, improving elderly patients’ ◦ Eating out evaluations of involvement may help to enhance ◦ Alcohol 31.5 THE DEVELOPMENT OF STRUCTURED EDUCATION: WHAT TO TEACH PEOPLE ABOUT DIABETES 435 ◦ How to modify insulin, depending on raised blood • Consequences of self-management decisions impact glucose level solely on the patient, their family and carers. ◦ Varying meal times and changes in routine • Acquiring new information is not easy. ◦ Importance of exercise DESMOND acknowledges that day-to-day, minute- In the UK, many groups have developed local ed- by-minute decisions (e.g. food choices, physical activ- ucation courses for type 2 diabetes, but only two ity, medication-taking, monitoring, etc.) which affect have been validated and published in peer review outcomes are made by the patient themselves. The per- journals: son with diabetes is thus responsible for managing their own condition • DESMOND (Diabetes Education and Self Man- agement, Ongoing and Newly Diagnosed) was de- The DESMOND randomized controlled trial veloped by a large collaborative of health care professionals and patients and designed to become Following a successful pilot study conducted in 17 a national programme. UK centres, the effectiveness of the DESMOND programme was investigated in a randomized controlled • X-PERT was designed in Burnley by Trudi Deakin, trial [14]. A total of 834 patients with newly diagnosed a specialist dietician and. following initial success. type 2 diabetes was recruited from 203 general prac- it is now being rolled out nationally. tices and randomized either to the control group, who received a total of 6 h of one-to-one education from a 31.5.1 DESMOND practice nurse, or to the DESMOND programme. DESMOND was set up to fill a gap identified by the The DESMOND curriculum, which was designed to National Institute for Health and Clinical Excellence be accessible to a wide range of people with diabetes, (NICE), who found very little evidence of effective was as follows: structured education for type 2 diabetes in the UK • Tell your story [13]. A large collaborative of psychologists, primary and secondary care physicians, specialist and practice • Professional story – normal physiology nurses, dieticians, educators, managers and people with • What is diabetes? diabetes met to address this void. Their vision was to develop a national educational programme, based on • Diet – done as games, choices patient-centred principles. To quote verbatim from the • Sugars informative website (www.desmond-project.org.uk/): • Fats “The educators or facilitators delivering the pro- • gramme are healthcare professionals working in the Monitoring community - mainly practice nurses, diabetes spe- • Physical activity cialist nurses or dieticians. Resources include patient support material especially written or produced for the • Complications programme and meeting its empowering philosophy. • Participants are not ‘taught’ in a formal way, but are Goal setting – be specific rather supported to discover and work out knowledge, The programme could be provided either as a sin- and to allow this to inform the goals and plans they gle 6-h session or in two 3-h sessions. The aim was make for themselves.” to incorporate it into routine diabetes care, delivered in the community. The DESMOND educators The principles of DESMOND received intensive training both in adult education techniques and facilitation of group learning, and adopted • Each individual is responsible for the day-to-day a non-didactic approach, so that learning was elicited management of their diabetes. rather than taught. Participants were encouraged to focus on lifestyle changes, such as food choices, physical • People make decisions to achieve best quality of life. activity, and on their personal risk factors. They were • Barriers to self-management lie in the individual’s then invited to choose a specific, achievable goal of personal world. behaviour change to work on. 436 CH 31 DIABETES EDUCATION IN THE ELDERLY

Measures of biomedical, lifestyle and psychosocial chart about people’s feelings about different aspects status – including depression – were made at baseline of diabetes. and 4, 8 and 12 months. All participants had recently Each person had a slightly different experience so discovered that they had type 2 diabetes, and therefore these things were coming up – there were things you made lifestyle changes in response to whatever advice had probably experienced yourself but hadn’t really they had received. As one might expect, this led thought about, and a picture developed of attitudes. to a marked fall in HbA level over the first 12 Towards the end of the course, the flip chart was 1c reproduced and we went through it all again. It was months in both groups, mirroring the findings of the interesting to find how people had changed. One lady UK Prospective Diabetes Study [15]. No significant when she came on the course said she felt like a leper, changes were found in biochemical measures, but there which I found surprising, but at the end of the course was a significant fall in body weight at all three time that had gone and she was feeling far more positive. = intervals in the intervention group (overall p 0.025), Someone would make a comment, which would even though weight reduction was not a specific target trigger something else, which without the previous of the programme. There was also an unexpected comment may not have come out. I think with things positive effect in the DESMOND group on abstinence like diabetes, you do need to talk about it. And to be from smoking after 6 and 12 months, with an odds able to do that with people in similar circumstances - ratio of 3.56. to hear what they feel and what their experience had Four distinct health beliefs about diabetes were mea- been, helped towards this positive attitude that came sured in the DESMOND trial, and all showed consis- out at the end of the course. tent and significant improvements (p < 0.001) in the I don’t believe that you would get the same level intervention group, who had a greater understanding of of information from your GP or practice nurse as their illness, including its seriousness and permanence you wouldn’t have the amount of time with them. To be able to have two half-day sessions with two with more confidence about their ability to affect the knowledgeable people that could answer questions course of their disease. The DESMOND group also = and talk about different issues was very reassuring had lower depression scores at 12 months (p 0.032), and supported everything you had learnt from your though there was no difference found in measures of practice nurse, your GP and websites and what have diabetes-specific emotional distress. you and I believe that everybody when they left that The DESMOND trial showed that a structured course felt very, very positive about it.” education programme for patients with newly diagnosed type 2 diabetes would lead to sustained 31.5.3 X-PERT programme benefits in health beliefs and depression, but did not demonstrate any difference in HbA1c level in the The X-PERT programme [16] was developed as first year. Thus, the programme could be introduced a patient-centred, group-based, self-management widely across primary care. programme based on theories of empowerment [6] The DESMOND educators observed that older pa- and discovery learning [17]. The effectiveness of tients contributed to the group and brought valuable the programme, as measured by clinical, lifestyle experience, but that they may have needed a different and psychosocial outcomes was demonstrated in a approach at times (see adaptations for the elderly). randomized controlled trial. For this, 157 participants were invited to attend 31.5.2 Case study six weekly 2-h sessions in groups of 16 people with diabetes and four to eight carers. The pro- Ken, aged 64, had recently been found to have gramme (see Table 31.4) aimed to develop skills type 2 diabetes and recorded his impressions of the and build confidence to enable people to make DESMOND programme: informed decisions about their diabetes self-care. “It was very interesting that a group of people who Separate sessions were held for Urdu-speaking were all recently diagnosed as type 2 diabetic were sit- participants, with a translator present. The 157 ting together as a group with two very knowledgeable controls were offered one-to-one sessions with a nurses who were talking to us about diabetes. These dietician (30 min), practice nurse (15 min) and a GP were very much group sessions and we were all given (10 min). The mean age of participants was 61 years the opportunity to talk about experiences and discuss in each group, and the mean duration of known issues and all sort of things were put up on a flip diabetes was 6.7 years. Some 14 months later, the 31.6 ADAPTATIONS OF EDUCATIONAL PRINCIPLES FOR THE ELDERLY 437

Table 31.4 The content of the X-PERT programme. Topic Description Week 1: What is Diabetes? Explore what happens to food when we eat it; self-monitoring of diabetes; diabetes treatments; feelings about living with diabetes. Dispel myths by using visual educational materials. Week 2: Weight Management Examine the ‘balance of good health’ model and use food models to distinguish between food containing protein, fat and carbohydrate. Inform about sensible eating whilst exploring barriers in doing so. Advise about the benefits of exercise and give practical examples including information about local exercise-on-prescription schemes. Week 3: Carbohydrate Perform a group task, developed to show the effect of quantity and quality of Awareness carbohydrate food on blood glucose levels. Use ping-pong ball models and laminated food pictures to dispel the myths surrounding glucose, sucrose and starch. Week 4: Supermarket Tour Address some common confusion surrounding dietary fat, sugar and food labelling. Encourage a diet that is enjoyable, variable and balanced whilst dispelling the concept of ‘good’ and ‘bad’ foods. Week 5: Complications & Discuss how to reduce the risk of developing longer-term complications through Prevention lifestyle changes, treatment and regular monitoring. Use visual educational aids to explore medical conditions in layman terms such as nephropathy, retinopathy, arteriosclerosis, neuropathy and blood pressure. Week 6: Evaluation & Play “Living with diabetes”, a board game to bring the X-PERT programme to a close Question time in a relaxed manner, reinforcing the main messages whilst encouraging participants to reflect on how much they have learnt. Goal Setting: Last 20 The final 20–30 minutes each week involves the goal setting component of the minutes each week empowerment model. Participants obtain and examine their health results, the implications of them and acceptable ranges. If participants make an informed decision to work on improving any of their health results, they work through the five step empowerment model. Psychosocial aspects of diabetes i.e. fitting diabetes into life rather than fitting life into living with diabetes. An important aspect of the empowerment model is to respect the decisions made by some of the participants not to goal-set. Patient Manual Resource manual given to participants at the beginning of the course. Background reading, health results and goal setting material added each week as appropriate. intervention group had improved control (HbA1c 31.5.4 Comment on educational content –0.6%versus + 0.1%, p<0.001) and a greater Local health care commissioners are responsible for reduction in body weight and waist circumference. funding educational and self-management programmes The X-PERT participants were less likely to have for people with diabetes. DESMOND and X-PERT increased their medication, and more likely to be are two examples of validated structured education exercising and performing footcare self-management. programmes which fulfil NICE guidelines, and have They were ‘much more satisfied’ with their diabetes subsequently been rolled out widely in the UK. Most treatment compared to those receiving individual elderly patients will benefit from this type of pro- appointments (p < 0.04), and had a higher total gramme, but if cognitive problems are suspected then empowerment score (p < 0.04), but there was any existing course must be modified along the lines no significant difference in their overall quality described in the following section. of life. Following the success of the trial, the X-PERT 31.6 Adaptations of educational programme now has a written curriculum, visual aids, principles for the elderly a ‘train the trainers’ course, an evaluation scheme and a quality assurance programme, and is being rolled out While the general principles for adult education are to other centres. well established, special consideration needs to be 438 CH 31 DIABETES EDUCATION IN THE ELDERLY given to elderly people and the factors which may • A reduced ADL (activities of daily living) ability. limit their ability to deal with new information and • A higher risk of hospitalisation in the previous year. the need for change. Older people are often aware that their ability to take in new ideas may be limited, and • An increased need for assistance in personal care. this may reduce their self-confidence. They should be encouraged to become actively involved in working The specific educational needs of the elderly have been out what they can do and, whenever possible, given addressed in a study conducted in Germany [19], where the opportunity to practice skills. Diabetes may be education has been an intrinsic part of diabetes care for perceived as a small and relatively unimportant part of many years. A group from Jena recognized that many their life and they may feel reluctant to make changes. elderly patients were not keeping up with the routine Many older people expect to be told what to do, and education programme (Treatment and Teaching Pro- find it disconcerting to be invited to make their own gramme; TTP), mainly because of impaired neuropsy- decisions. chological function. Thus, a programme was devised Information needs to be personalized, with the pace for elderly patients with poor cognitive function, which of delivery and specificity adapted for the individual. took account of their reduced life expectancy. The aim Language should be basic but not patronizing, with was to enhance the patient’s quality of life by main- the information selected so that people learn what they taining autonomy and independence. The course was need to know, rather than everything there is to know designed to emphasize individual therapeutic goals and about diabetes. Avoid trying to put over more than one to highlight catabolic symptoms, acute complications concept at a time. It is important to establish the meal and depression. A total of 102 patients was included in pattern and the extent to which the person has control the study, and cognitive function was assessed by mul- over what they eat before exploring options for change. tiple tests. Sixty-eight patients, who were found to have Stereotypes of the elderly are usually negative, cognitive impairment (IQ score <91), were random- and this in itself may reduce self-confidence and ized to take part either in a standard TTP programme or hence the learning ability of the older person. Any in the geriatric DICOF TTP. This adaptation places less assessment must try to overcome the negative, while emphasis on theoretical knowledge, such as pathophys- being realistic. iology or blood glucose self-monitoring and detailed carbohydrate counting, and leaves room for intensive 31.6.1 Do the elderly have special needs in training in the practical matters of insulin injection, diabetes education? coping with hypoglycaemia and urine glucose testing. A key feature of the programme is repetition to re- There is an assumption that old people will have spe- inforce key topics. The DICOF TTP takes 5 days to cial educational needs, in particular concerning dif- complete, and there are nine education units: ficulties with understanding and in concentrating on abstract concepts such as the pathophysiology of di- • Unit 1: What is diabetes? abetes. In addition, people with diabetes could be at • Unit 2: Urine/blood glucose self-monitoring greater risk of cognitive impairment than those without diabetes, which could affect their ability to care • Unit 3: Insulin injection – group and one-to-one ed- for their diabetes. These questions were addressed by ucation Sinclair et al. [18], who studied 369 patients with di- • Units 4, 6 and 7: Nutrition abetes over the age of 65 years and compared their cognitive function with a carefully matched group of • Unit 5: Hypoglycaemia non-diabetics. The odds ratios (95% CI) for normal • Units 8 and 9: Late complications – group and cognitive test results in subjects with diabetes, after one-to-one education adjusting for all significant variables, was 0.74. Dia- betic subjects with a Mini Mental State Examination Each session is facilitated by a trained educator and an (MMSE) score <23 had the following associations (all interested physician. with p < 0.001): At the follow-up examination 6 months later, there were no differences in HbA levels, but the DICOF • A reduced involvement in diabetes self-care. 1c group showed improvements over the standard group • Less diabetes monitoring. in the following areas: 31.8 SPECIFIC CASES 439 • better self-management (insulin injections and The assessor should decide whether the elderly self-monitoring) person would benefit from group education, perhaps with specific additional support, whether they need • greater satisfaction with the education programme special treatment such as one-to-one education, or • higher quality of life whether the education needs to be provided to carers rather than to the elderly patient themselves. • reduced fear of hypoglycaemia. Thus, although it is difficult to produce hard evidence for customized education programmes for the elderly, 31.8 Specific cases the Jena group have proved the effectiveness of their carefully designed education package. 31.8.1 Insulin refusal Many people of all ages find the prospect of taking 31.7 Assessment of the educational insulin daunting and have to be convinced that it is re- needs of the individual ally necessary. This is particularly true if the reason for recommending insulin is to achieve an HbA1c target in an asymptomatic patient, rather than to relieve symp- 31.7.1 Timing of assessment toms of hyperglycaemia. The major barriers to taking insulin are the fear of self-injecting and anxiety about • The first health care professional (HCP) to speak hypoglycaemia. Most people have heard of someone with the newly diagnosed person is the best placed who has terrible trouble with ‘hypos’, and assume that to make an initial assessment of their response to the it is an inevitable consequence of taking insulin. news of the diagnosis and their ability to take in new Elderly people are often reluctant to accept insulin information. This should include an assessment of injections, and feel it is should not be necessary in physical (e.g. vision, hearing) and cognitive function someone of their age. They are less convinced by the and potential to work within a group. arguments of long-term risk reduction and are not dis- posed to make changes unless they see a clear and 31.7.2 Nature of assessment immediate benefit, such as a relief of the symptoms of hyperglycaemia. It is incumbent on health care profes- The assessment should be considered from two per- sionals to try to assess the actual benefit of insulin in spectives: the context of comorbidities and life expectancy before 1. The HCP assessment of patient and carers: recommending a change. • Physical: restrictions which may impact on ability Exploration of the individual barriers to insulin and to work within a group (e.g.. deafness, reduced the reasons for refusal should help identify solutions, vision, cognition, mobility). which could be: • Psychological: factors which may impair ability • intensified educational support during the initiation to take in new information (e.g. depression, low stages self-confidence, level of independence, flexibility, passivity). • administration by district nurse or a carer, either • Social: factors which may limit ability to meet temporarily or in the long term; and/or targets (e.g. isolation, exercise capacity, realistic • the selection of a suitable insulin delivery device. targets for control). In some circumstances the person may remain uncon- 2. The patient’s (and their carer’s) assessment of the vinced and decide to accept the risk associated with process: poor glycaemic control. • Their expectations of what the system may provide. • Preconceived ideas and beliefs, readiness to make 31.8.2 Choice of insulin changes. The insulin regimen should be discussed and chosen in • Perception of their mortality/morbidity. the context of the ability of the elderly person to make • Physical restrictions. day-to-day decisions about dose and activity. Those 440 CH 31 DIABETES EDUCATION IN THE ELDERLY with a regular routine of food intake and activity may Elderly people with longstanding type 1 diabetes can prefer a twice-daily mixture which can be adjusted to be a particularly difficult group. They have often kept suit a regular meal pattern. For people with erratic tight control of their blood glucose levels over many eating habits or variable activity levels, a decision years, which is why they have lived to a great age. should be made about the individual’s ability to vary They may accept hypoglycaemia as an inevitable con- the insulin dose. Although, in general, elderly people sequence of tight control. However, increasing frailty may not take well to the decision making associated and forgetfulness may make decisions about food and with a basal bolus regimen, some are happy to take insulin more erratic, and it is not uncommon to see it on. Those who cannot make their own decisions, big swings in the blood glucose level, with the po- but who may not be able to keep to a regular routine, tential for severe hypoglycaemia. This can be very are better treated with a simple insulin regimen such difficult to correct, as the strategy of avoiding high as once- or twice-daily background insulin. In this and low levels by use of intermediate and long-acting situation, the aim should be to control symptomatic insulin does not work well in this insulin-deficient hyperglycaemia and avoid hypos, rather than to achieve group. Written advice on insulin adjustment, depend- a specific HbA1c target. ing on the blood glucose level may help. Carers may For those unfamiliar with blood glucose monitoring be educated to support decision making, but the re- there is even more to learn, and this technique should sults are often unsatisfactory, particularly for people in be taught before insulin is introduced, unless the need institutional care. for insulin is urgent. The pace of the education should be dictated by the individual, with repetition and rein- 31.8.4 Coping with illness forcement of information [18]. Insulin administration devices should be chosen after ‘Sick day’ rules can be complicated, and elderly people an assessment of the patient’s dexterity and visual often become confused as a consequence of intercur- impairment. The Innolet device is particularly useful rent infection and so may not be able to implement for those with poor vision or reduced dexterity, but advice. Dehydration can lead to a rapid deterioration in unfortunately this may limit the choice of insulin as renal function, and this carries a risk of life-threatening only Mixtard 30, Insulatard and Detemir are available lactic acidosis for those taking metformin. As an acute in this format. intercurrent illness (e.g. gastroenteritis) is likely to be an infrequent event and verbal advice may have been 31.8.3 Hypoglycaemia forgotten, written guidelines should be provided for both patients and carers. These should be simple and Elderly people are at particular risk from hypogly- should recommend frequent monitoring of the blood caemia as the consequences are potentially more glucose during the illness, seeking medical advice if severe, with increased vulnerability to confusion, the results are outside a specified range. Specific indi- cognitive impairment and neurological deficit. cations for urgent medical treatment, such as vomiting, Hypoglycaemic episodes may be acute and severe, breathlessness, confusion or altered conscious level, sometimes presenting as a stroke, but they may should be listed. also be mild and recurrent. The signs of mild hypoglycaemia can be more difficult to recognize, 31.8.5 Polypharmacy as the only manifestation may be confusion and cognitive impairment, which is often attributed to Many people with diabetes are now advised to take other causes. Although most people are alert to the large numbers of medications to reduce cardiovascular possibility of hypoglycaemia in somebody taking risk. There is evidence to demonstrate that people are insulin, sulphonylureas also carry a significant risk more likely to adhere to treatment if they understand of hypoglycaemia, particularly if their food intake and accept the reasons for taking it, and this underpins is reduced or their renal function deteriorates. Many the self-management philosophy of programmes such elderly people receiving oral therapy will not be mon- as DESMOND and X-PERT. Complicated regimens itoring their blood glucose, and the problem may only may be confusing to elderly people, although there come to light when the person is admitted to hospital are various practical ways of simplifying things, for with cognitive or neurological impairment and blood example daily dosing boxes. The role of the local glucose monitoring reveals persistently low levels. pharmacist is crucial to the success of such methods. 31.10 ACKNOWLEDGMENTS 441 31.9 Conclusions 8. Stewart M and Roter D. (1989) Communicating with medical patients. Sage, Newbury Park, CA. 9. Garcia R and Suarez R. (1997) Diabetes education in The general principles of diabetes education for adults the elderly: a 5-year follow-up of an interactive ap- apply equally to the elderly population. However, the proach. Patient Education and Counseling 29: 87–97. specific needs of the elderly – taking into account 10. Kemper S. (2003) Elderspeak. Acoustical Society of attitudes and disabilities – need to be assessed so that America Journal 113: 2295–5. modifications may be made to meet their individual 11. Williams K, Kemper S and Hummert ML. (2004) needs. Enhancing communication with older adults: Overcom- ing Elderspeak. Journal of Gerontological Nursing, 30: 1–8. 31.10 Acknowledgments 12. Wensing M, Wetzel S, Hemsen J and Baker R (2007) Do elderly patients feel more enabled if they had been The authors thank many experienced diabetes educa- actively involved in primary care consultations? Patient tors in the planning of this chapter, and in particu- Education and Counseling 68: 265–9. lar Penny Meade, the community diabetes facilitator 13. Skinner TC, Carey ME, Cradock S et al. (2006) Di- at Northampton Diabetes Centre, and Kathryn Sutton, abetes education and self-management for ongoing senior dietician at the Ipswich Diabetes Centre, who and newly diagnosed (DESMOND): Process modelling provided thoughtful and concrete suggestions. of pilot study. Patient Education and Counseling 64: 369–77. 14. Davies MJ, Heller S, Skinner TC et al. (2008) Effec- References tiveness of the diabetes education and self management for ongoing and newly diagnosed (DESMOND) programme for people with newly diagnosed type 2 1. National Institute for Health and Clinical Excellence diabetes: cluster randomised controlled trial. Br Med J (NICE) (2003) Guidance on the use of patient- 336: 491–5. education models for diabetes. Technology Appraisal 15. United Kingdom Prospective Diabetes Study (UKPDS) 60. NICE, London. Group. (1995) UKPDS 13: relative efficacy of ran- 2. Structured Patient Education in Diabetes – Report from domly allocated diet, sulphonylurea, insulin or met- the Patient Education Working Group. (2005) De- formin in patients with newly diagnosed non-insulin partment of Health and Diabetes UK, London. DH dependent diabetes followed for three years. Br Med J 4113197. 310: 83–8. 3. Trento M, Passero P, Borgo E et al. (2004) A 5-year 16. Deakin TA, Cade JE, Williams R and Greenwood randomized controlled study of learning, problem solv- DC (2006) Structured patient education: the Diabetes ing ability, and quality of life modifications in people X-Pert programme makes a difference. Diab Med 23: with type 2 diabetes managed by group care. Diabetes 944–54. Care 27: 670–5. 17. Brunner J. (1966) Toward a Theory of Instruction. 4. Kulzer B, Hermanns N, Reinecker H et al. (2007) Ef- Harvard University Press, Cambridge, MA. fects of self-management training in Type 2 diabetes: a 18. Sinclair AJ, Girling AJ and Bayer AJ (2000) Cognitive randomised, prospective trial. Diabet Med 24: 415–23. dysfunction in older subjects with diabetes mellitus: 5. National Service Framework for Diabetes: Stan- impact of diabetes self-management and use of care dards. (2001) Department of Health, London services. All Wales Research in Elderly (AWARE) (www.doh.gov.uk/nsf/diabetes). Study. Diabetes Res Clin Pract 50: 203–12. 6. Anderson RM and Funnell MM. (2005) The Art of 19. Braun, U., Muller, A., Muller, R. et al. (2004) Struc- Empowerment, 2nd edition. American Diabetes Asso- tured treatment and teaching of patients with Type 2 ciation, Alexandria, USA. diabetes mellitus and impaired cognitive function – the 7. Care Planning in Diabetes. (2006) Report from DICOF trial. Diabetic Medicine 21: 999–1006. the joint Department of Health and Diabetes UK Care Planning Working Group. www.diabetes.nhs. uk/downloads/care planning in diabetes report.pdf 32 Supporting the Family and Informal Carers

Antony Bayer1 andAlanSinclair2 1Department of Geriatric Medicine, Academic Centre, Cardiff University, Llandough Hospital, Vale of Glamorgan, UK 2Bedfordshire and Hertfordshire Postgraduate Medical School, University of Bedfordshire, Luton,UK

its complications and in adopting and maintaining a Key messages healthy lifestyle. There is a signiﬁcant literature on the role of family • Informal carers (caregivers) are the primary carers in childhood (type 1) diabetes [1, 2], where it is source of everyday advice, emotional support taken for granted that parents will play an active role and practical help for many older people with in positively inﬂuencing management and make efforts diabetes. to ensure that glucose control is optimal. Numerous • A supportive environment for carers must be studies have examined the value of psycho-educational provided that allows them to receive educational interventions and family support on health outcomes advice about diabetes and its management, and creates opportunities for greater involvement in of children and young people with insulin-dependent treatment decisions. diabetes, and have shown that family-based interven- • The effects of caring on the carer in terms tions can be effective in improving diabetes knowl- of physical, emotional, social and economic edge, promoting adherence to management regimens well-being must not be overlooked and achieving better diabetic control and increasing quality of life and emotional health [3, 4]. In contrast to this active research and clinical inter- 32.1 Introduction est in carers of young people with diabetes, there is much less acknowledgement from health profession- Most people with diabetes live at home with support als and policy makers of the important role of family from family and friends. Many of these patients would and informal carers in delivering and monitoring op- not regard themselves as requiring care, and most timal diabetes care for older adults. In the UK, the close relatives would not regard themselves as carers. Diabetes National Service Framework (NSF) [5] em- Nevertheless, in reality, nearly all of the 2 million or so phasizes the importance of empowering people to con- older people with diabetes in the United Kingdom will trol day-to-day management of their diabetes and does look to their family and friends for advice and practical mention that “...where appropriate, ...carers should and emotional support in managing their diabetes and be fully engaged in this process”. Carers are also

Diabetes in Old Age Third Edition Edited by Alan J. Sinclair © 2009 John Wiley & Sons Ltd. ISBN: 978-0-470-06562-4 444 CH 32 SUPPORTING THE FAMILY AND INFORMAL CARERS mentioned as needing to be alert to the dangers of aspects of diabetes-specific care [9]. In the present diabetic emergencies, but no consideration is given to authors’ study of older patients attending hospital- how this should be achieved or that caring relatives based diabetes clinics – about half of whom were and friends may have needs and wants of their own. taking insulin therapy – nearly all reported that they This does not seem to reflect adequately the qualita- regularly received help with day-to-day activities or tive research that was undertaken to inform the devel- looking after from someone else [10]. Thus, informal opment of the NSF [6]. Whilst only carers of children carers are the main source of support for older people were included in the focus groups and in-depth inter- with diabetes, helping with emotional adjustment to views, the older patients who were interviewed stated the diagnosis, adherence to diabetes regimens, manage- that they wanted partners and carers to be with them at ment of complications and coping with everyday life the time they were told their diagnosis – both for emo- [11–13]. Yet, service providers too often ignore the po- tional support and to be able to learn together about the tential needs of the carer because the focus of concern condition and how to deal with it. They also suggested is inevitably the needs of the care recipient [14]. that partners or other carers should be given more opportunities for learning about diabetes. This was not only because carers required the information to help 32.2 Who are the carers? look after the person, but also because the adjustments related to living with diabetes affected both those with Informal carers may be family members (e.g., spouse, diabetes and the people with whom they share their sibling, child or grandchild), friends, neighbours, or lives. It was therefore suggested that carers should be members of a specific support system such as a church invited to attend routine and other appointments with or social organization. The results of the carers module the service user and to ask questions, as required. of the General Household Survey (GHS) of 2000 [15] It is encouraging that the updated guideline on man- reported that just over half (52%) of carers were agement of type 2 diabetes from the National Institute looking after a parent (or parent-in-law) and 18% for Health and Clinical Excellence [7] acknowledges were caring for their spouse or partner. Two-thirds that diabetes is “...predominantly managed by the per- were caring for women, reflecting the predominance son with the diabetes and/or by their carer as part of of women in the older age groups. The most common their daily life”, and that, if the patient agrees, fam- age group of carers was between 45 and 64 years ilies and carers should have the opportunity to be (24%), although a significant proportion (12%) were involved in decisions about treatment and care and themselves aged over 65 years. they should also be given the information and sup- Whilst there have been few reports that have focused port that they need. A key priority for implemen- specifically on the informal carers of older people with tation is that structured education should be offered diabetes, in general these seem to reflect the findings “...to every person with diabetes and/or their carer of the broader carer literature [9, 16, 17]. at and around the time of diagnosis and with rein- Generally, one person takes on the lead carer role, forcement and review on an annual basis”. Yet, the with a secondary network of support provided by UK National Director for Diabetes published a doc- more distant relatives and friends. The role of primary ument in 2007 that claims to “...set out how ser- carer is decided by generation (when available, spouses vices are changing to meet the needs of patients and whether husband or wife take on most care), gender how they need to change in the future” [8]. This (typically the hierarchy of selection is wife, daugh- states that the average patient with diabetes will spend ter, daughter-in-law and even granddaughter before three hours a year with a professional and the remain- sons), geography (those living with, or closest to the ing 8757 hours caring for themselves. That family person) and competing responsibilities (employment, or friends might play any role at all is not consid- childcare and other family responsibilities). Widow- ered. hood becomes more common with increasing age, and In practice, the involvement of non-professionals in so children inevitably then become more actively in- various aspects of care of older people with diabetes volved. Husbands rarely give direct help to wives who is commonplace. In an observational study of patients are carers, and women are more likely to have to give aged ≥70 years participating in a study of type 2 up their job to take on caring responsibilities [18, 19]. diabetes in primary care, between 22% and 50% of It is fairly clear that in many cases there is what is their family members reported helping with various described as a ‘caregiving trajectory’, where the nature 32.3 WHAT DO CARERS DO? 445 of caregiving varies over time [20] for most adults. 32.3 What do carers do? This may be especially true in poorer communities with three phases usually identified: semi-care, care, The term ‘carer’ relates to a broad spectrum of tasks, and dying. ranging from emotional support, to organizing help, Ethnicity also plays an important role, with to providing company, to doing household tasks and differences in the caring roles and expectations [21]. to help with intimate personal care. In the GHS [15], Due to patterns of migration, there are proportionally over two-thirds (71%) of carers provided practical more older men requiring care within ethnic minorities help such as preparing a meal, shopping and doing and there is less acceptability of available formal laundry, 60% kept an eye on the person being cared care services and institutional care. Consequently, for, and 55% provided company. Smaller proportions of carers provided more intimate forms of help. About there is a greater care burden on families. In their one-quarter (26%) gave assistance with personal care study of South Asian carers, Townsend and Godfrey such as washing, 22% administered medicines, and [22] found that caring is influenced by gender role 35% provided physical help, for example with walking. stereotypes and filial responsibility. Spouse carers The availability of a close family member or friend were less common and carers tended to be younger to take on the role of carer, rather than need, seems [23]. Sons took a more active responsibility for to be the most important predictor of having someone decision making, organizing care and assisting with involved in medical care [27], and not all recipients of instrumental activities of daily living, with daughters care appreciate the level of involvement (too little or and daughters-in-law helping with personal care and too much) that is provided [28]. housework. Often, daughters and daughters in law The general carer research literature distinguishes did not regard themselves as carers, but the support between caring (the affective component) and given to their relative was considered part of normal caregiving (the behavioural component) [29]. ‘Care family responsibilities. The belief that ethnic minority providers’ give hands-on-care (e.g. dressing, bathing, daily supervision, cooking, managing finances, families ’look after their own’ and so do not require transportation) and ‘care managers’ arrange for others attention is not supported by evidence [24]. to provide care (e.g. organizing a nurse to attend daily In many instances, adults within ethnic minority to give medication or dress leg ulcers, a professional groups may be both patients with diabetes and have carer to attend to personal care and provision of meals, a caring role for younger people. For instance, in a an accountant to manage finances and social care study of 109 urban midlife African American women assistants to provide companionship and supervision with type 2 diabetes, 60% were grandmothers who during the day). Care providers are most often spouses had higher levels of diabetes-related emotional stress (especially wives) and daughters/daughters-in-law, and worse glycaemic control than those without grand- whilst care managers are most often husbands, adult mother status, and yet had higher quality of life scores children (especially sons), friends and other relatives. than non-grandmothers [25]. Care providers tend to be more burdened/stressed than The contribution of children to care should not be care managers, but inevitably have less contact with overlooked. In a study of child carers of adults with professionals who might be able to provide practical diabetes by Jacobson and Wood [26], one in five and emotional support [30]. A study conducted by Murphy et al. [16] in a fam- were looking after grandparents, having begun caring ily practice in the United States looked specifically at at a mean age of 11 years. Most provided care at the supportive family members of adults with type 2 least several times a week, ranging from calling to diabetes (mean age 59 years) and identified two broad check on the adult or staying with them overnight, to categories of family participation. As well as the con- performing glucose testing and giving medication and ventional supportive family member (primary carer or insulin injections. The youngest child administering helper) who provided supportive tasks in the care of the insulin was aged 5 years. Nearly half of the children illness, there was often the ‘family health monitor’, an had no education about diabetes care, not even from internal health expert usually consulted before any con- the family member with diabetes. sultation with external resources, including the doctor. 446 CH 32 SUPPORTING THE FAMILY AND INFORMAL CARERS

Other reported help given was ‘keeping enough medication Blood sugar 6% on hand’ and ‘following a diet’. Between 23% and 38% monitoring 9% of family members also reported participating regularly in the patient’s medical encounters, whilst 20–40% General support Help with diet rarely or never discussed diabetes-related issues with 15% 48% the doctor. When they did go to appointments, family members usually talked to the doctor with the patient present, although the most common reason for want- Help with ing to talk was to get their own questions answered. medication Prognosis was discussed less frequently with family 22% members than were test results, treatment issues and Figure 32.1 Helping activities provided by family preventive strategies. Carer needs were almost never members to adults with type 2 diabetes. (Adapted from considered. Predictors of participation in the patient’s Ref. [16]). medical encounters included older age and a greater physical impairment of patients and increased involve- This person is described as fulfilling a unique execu- ment in diabetes-related and general care. tive function as an authoritative information resource Among carers of community-living elderly patients and supervisor – acting to critically evaluate medical advice before family members incorporate the informa- attending hospital-based diabetes clinics in Birming- tion into daily practice. Three-quarters of the patients ham [10], up to 90% reported providing help with in- identified such an individual within their family, and strumental activities of daily living, such as shopping, often this was not the same person as the primary housework, preparing meals, finances and transport, carer. No relationship was shown between the pres- and up to 25% help with personal care such as washing ence or absence of a family health monitor and the and bathing/showering, walking about outside, dress- level of metabolic control as measured by glycosolated ing/undressing, getting in and out of bed and toileting. haemoglobin (HbA1c) level. This suggests a need for As would be expected, when patients are less phys- health professionals to recognize and involve family ically functional, when family members are spouses, health monitors in the therapeutic team, so that they when they provide more assistance with basic care, may impact more positively on management. and when they have a greater understanding of dia- The study by Murphy et al. also identified the most betes management issues, family members are more frequent helping tasks undertaken by the primary carer (Figure 32.1). By far the most common activity was likely to provide assistance with diabetes-related care helping the person with diabetes with their diet. This [9]. Cognitive impairment is also a strong predictor of a included food selection and preparation (‘helps buy greater need for the involvement of carers in supervis- the right foods’, ‘cooks properly’), reminders about ing medication, monitoring blood glucose and helping proper diet (‘keeps after me about diet’, ‘watches with personal care [31]. diet’) and support for dietary restrictions (‘cooperates Using data from the Oldest Old Study, a nationally at meals’, hides sugar’). Help with medications was representative survey of people aged ≥70 years in the both general (‘keeps track of medications’) and specific US, Langa et al. [32] determined the weekly hours of (‘buys medicine’, ‘reads directions on medicine bot- informal caregiving received by community-dwelling tles’). General support was defined as ‘encouragement’ elderly individuals with and without a diagnosis of or ‘talking to me’. Other helping activities included diabetes. Those without diabetes received an average financial support, reminders about medical appoint- of 6.1 h of informal care, those with diabetes taking ments, assistance with hygiene and exercise. In the primary care-based study conducted by Sil- no medications received 10.5 h, those with diabetes liman et al. [9], it was reported that between 6% and taking oral medications received 10.1 h, and those on 17% of older patients with type 2 diabetes received reg- insulin received 14.4 h of care (p<0.01). Disabilities ular help from family members with basic activities of related to heart disease, stroke and visual impairment daily living, and between 37% and 48% with instru- were important predictors of a need for diabetes-related mental activities of daily living. The most commonly informal care. 32.4 WHAT EFFECT CAN CARING HAVE ON THE CARER? 447 32.4 What effect can caring have on that used the SF-36 questionnaire to measure aspects of health-related quality of life, found that the carers had the carer? a poorer mental but better physical well-being than the population norm [44]. A study of family caregivers of The responsibilities of the caring role can take their diabetes patients in Sudan, using the WHO 26-item toll on the physical, emotional, social and economic quality of life measure, found that those who were well-being of the family and others closely involved younger, single, less-educated and caring for people in care provision [33–36]. Most published studies of with more recently diagnosed illness, were relatively the experience of significant others of persons with vulnerable to the negative effects of caring [45]. This diabetes have been about adults with Type 1 diabetes. latter research group also published evidence that there These show that relatives may harbour even higher was greater concordance between the impressions of levels of concern and worry about the illness and its family caregivers on the patient’s quality of life in effects than the patients themselves [37], with spouses type 1 than type 2 diabetes, presumably because in reporting marital conflicts about diabetes management those with type 1 diabetes it was easier to define the and disturbed sleep [38]. Furthermore, it has been factors that adversely impacted on a patient’s quality reported that the diabetes self-management behaviour of life, such as diminished sexual desire or additional of husbands often deteriorated when conflict existed medical conditions [46]. with their spouses [39]. Patients with complications of their diabetes are A qualitative study conducted Stodberg et al. [40] likely to require more care, and this will place an of the ‘lived experience’ of being a close relative even greater burden on their carers. A small qualita- (significant other) of persons with type 1 diabetes tive study found that the development of diabetic foot identified four major themes: (i) living in concern ulcers leads to both patients and carers experiencing a about the other’s health; (ii) striving to be involved; reduction in social activities, increased family tensions, (iii) experiencing confidence; and (iv) handling the lost time from work and a negative effect on general illness. Many of the carers said that they lived a health [47]. Another study of patients with diabetic normal life and had come to accept diabetes as a foot ulcers (mean age 60 years) and their carers from normal part of life. At the same time, they felt they centres in the USA, UK and Europe found that patient needed to be constantly attentive to how the person and carer scores on the SF-36 were closely correlated, with diabetes was feeling, and lived their life waiting with healing associated with a large improvement in for the complications to come. They felt sorrow when the subscale related to emotional difficulties of the they watched the health of the patient deteriorate and, carers [48]. Starting insulin treatment for selected el- whilst they had found ways to handle the illness, many derly people with type 2 diabetes with poor glycaemic felt that they lacked adequate recognition and support control on tablet therapy improved not only the pa- from health care staff. When professionals took little tients’ quality of life and mood, but also limited carer notice of the significant others, carers felt humiliated strain, as measured by the general health questionnaire and neglected, and uncertain as to how best they should [49]. care for the ill person. They had questions, but no one A UK study estimated the annual average financial to direct them to. cost to working-age carers of looking after someone The emotional burden of caring is a recurring with type 2 diabetes to be £1300, but when earnings topic in the caregiving literature [41]. Levels of low were actually lost the cost was almost £11000. Carers mood and anxiety and rates of likely depression who lose earnings report higher levels of strain. Only (21%) were as high, or even higher, in partners of one-third of carers reported receiving state benefits, European-American and Latino patients with type 2 and the shortfall between earnings lost and benefits diabetes as they were in the patients, especially if received was substantial [50]. the partner was female [42]. Psychological distress in Not surprisingly, given that people who spend a lot either partner either increases or is positively corre- of time together will share many lifestyle behaviours, lated with marital discord, hostility and conflict which, non-genetically related partners of people with diabetes in turn, decreases disease-related problem solving and also have a greater than twofold increased risk of marital satisfaction and can affect disease management being diagnosed with diabetes themselves during their and disease progression [43]. A study in Taiwan of the lifetime compared to controls, and one in five display primary carers of elderly people with type 2 diabetes evidence of glucose intolerance [51]. 448 CH 32 SUPPORTING THE FAMILY AND INFORMAL CARERS While there is increasing support for a link between become partners in the provision of care, and supports caregiver burden and diabetes, this requires further test- them in best helping the person they are caring for. ing. For example, in a longitudinal study of frail elderly Conversely, without information, carers are more likely subjects in Japan, which included use of the Japanese to suffer from stress and consequently be less able to version of the Zarit Caregiver Burden Interview, di- continue to care. abetes did not appear to feature as an independent In Hennessey’s study there was a perceived lack of predictor of caregiver burden [52]. information about the nature and the expected course of diabetes (especially in those with comorbid condi- 32.5 What do carers want? tions), trepidation in handling tasks such as postoperative amputation care or coping with dialysis machines, Families and friends involved in caregiving speak of and fears concerning the occurrence of a diabetic crisis. their need for recognition, information and advice to All of the carers emphasized the importance of developing and implementing efficient caregiving routines help them in their caring role, and adequate support and mastering care techniques for successful diabetes services and respite when needed (Table 32.1). While care management. They looked for expert guidance and many carers are eligible to receive formal care services, support on how this might be best achieved, and the practical help from the family is often the preferred implementation of diabetes education programmes tar- option. This is not due to dissatisfaction with formal geted at family caregivers was strongly recommended. services, but rather there is a general sense that while Although any information provided must include informal networks exist an atmosphere of normality medical management of the disease, it is equally impor- can prevail [14]. tant that consideration be given to the social and prac- A qualitative study by Hennessy et al. [17] of fam- tical management of diabetes within a family context. ily members caring for elderly American Indians with Thus, diabetes education programmes should also offer diabetes investigated diabetes care management, the content on predictable psychosocial and behavioural challenges faced and the support services needed. The problems encountered in diabetes care management focus group participants reported a number of con- with older adults, how these problems can be addressed cerns, including anxiety about home care management, within the family, and where help is available when coping with psychosocial issues such as depression or family efforts have not been successful. Despite the non-compliance, decision making and communication increasing reliance on the internet for information pro- with other family members. The findings would seem vision, there still is little material on diabetes that is relevant to most informal carers trying to help and sup- specifically targeted at carers. port older people with diabetes in the community. The family members in the study conducted by A need for more information is the most common Hennessy and colleagues also highlighted perceived request of carers [53]. Subject to the consent of the gaps in the provision and continuity of formal care older person, carers want timely education and advice services, with carers stating that they often felt stranded about the specific health problems of the person they without sufficient professional backup for care. One are caring for, what they can do to help, and the common complaint was a difficulty in knowing who services available. Good information enables carers to to contact and of obtaining a prompt response. Whilst Table 32.1 What do carers want? the appointment of a case-manager should, in theory, address this, some artificial boundary or restriction is • Respect and recognition as a partner in care all too-often cited as a reason to be unable to help, • Timely explanation and relevant information or carers are told that ‘the case has been closed’ • The right skills and expertise to manage and care and they must go back to the end of the queue. The lack of an adequate response from professionals is • Knowing the options and what help is available short-sighted, as problems may then escalate and lead • Practical support, especially the opportunity to take a to carer breakdown and avoidable hospital admission, break or a need for permanent nursing home care. Levine • Appropriate and flexible services, available when they [54] described her sense of isolation and frustration are wanted with formal care providers in a personal account, • Adequate income entitled ‘The Loneliness of the Long-Term Caregiver’, where she said that it often felt as if “...she was 32.6 WHAT ARE THE BENEFITS OF CARER INTERVENTION? 449 challenging Goliath with a tiny pebble. More often than Table 32.2 Potential benefits of networking and sup- not, Goliath just puts me on hold.” ported informal care. Everyday problems such as substandard housing, • Provision of additional social and emotional support a lack of modern conveniences, lack of financial resources and reliance on others for transportation will • Increased availability and access to relevant health care exacerbate the burden on carers, as well as interfering information with the ability to develop a routine for the cared-for • Sharing of Good Clinical Practice person. A perceived absence of professional guidance • Promotion of improved health care behaviours among or support in dealing with psychosocial problems will both patients and carers mean that carers have to devise their own strategies • Increased mobilization of community-based diabetes re- for dealing with behavioural or psychological aspects sources of care, such as attempting to coax, cajole, or coerce • patients into compliance with care regimens [17]. Provide better opportunities for integrated diabetes edu- The importance of coordinating the activities of all cation family carers who provide assistance with diabetes care • Promotes leadership qualities is also emphasized, with primary carers expressing • Improved blood glucose management frustration when they are unable to inform and syn- chronize the caregiving efforts of those involved. Peri- odically holding a family meeting with or without the However, only three of the studies involved spouses, participation of healthcare providers can be as an ef- family or friends, and the mean age of the patients fective intervention to resolve or significantly improve involved was only 59.3 years. During a 6-week edu- understanding of diabetes care requirements [17]. cational programme with older patients with diabetes, those with participating spouses, compared to those without, showed a greater improvement in diabetes 32.6 What are the benefits of carer knowledge and metabolic control [64]. In contrast, the intervention? participation of family and friends in diabetes education group sessions for Native Americans had no effect In patients with chronic disease, adherence to medical on metabolic control in women with type 2 diabetes treatment is increased when family relationships are [65]. Indeed, social support may have different effects supportive [55], and this can improve outcomes [56]. for men and women. In the study conducted by Wing Certainly, patients who feel supported and ‘cared et al. [66], support from the spouse (in the same educa- for’ report a greater sense of well-being [57], fewer tional programme) acted positively on weight loss for depressive symptoms [58] and better general health obese women with type 2 diabetes, while participating [59]. However, family members seem to view diabetes without the spouse worked out better for men. as a more serious illness than those with the condition In the observational study by Silliman [9], patients [60], and this lack of concordance can lead to conflict. receiving more assistance from family members Certainly, if patients perceive care as controlling and were more likely to report that they were taking limiting it may have a negative impact, conceptualized their medications as prescribed, that they were as ‘miscarried helping’ in relation to adolescents with following their diabetes diets, and that there was some diabetes and their parents [61]. It is important that correlation between family member assistance and any carer input does not undermine patient autonomy random glucose levels. [13, 62]. There may be potential benefits of closer networking There is some (albeit limited) research evidence that and greater support for carers, and these have been the involvement of family and friends in diabetes care listed in Table 32.2. can improve metabolic control and the management of complications. A systematic review of prospective in- References tervention trials of social support on health outcomes in primary and outpatient care for type 2 diabetes iden- 1. Burroughs, T.E., Harris, M.A., Pontious, S.L., et al. tified six trials of adequate quality for review [63]. (1997) Research on social support in adolescents with Most carried evidence in support of the idea that so- IDDM: a critical review. The Diabetes Educator, 23, cial support is influential on self-care and outcomes. 438–48. 450 CH 32 SUPPORTING THE FAMILY AND INFORMAL CARERS

2. Lowes, L. and Lyne, P. (1999) A normal lifestyle: American Indian elders. The Diabetes Educator, 25, parental stress and coping in childhood diabetes. British 747–54. Journal of Nursing, 8, 133–9. 18. Parker, G. (1985) With due care and attention: a review 3. Armour, T.A., Norris, S.L., Jack Jr, L., et al (2005) of research on informal care. Family Policy Studies The effectiveness of family interventions in people Centre, London. with diabetes mellitus: a systematic review. Diabetic 19. Twigg, J., Atkins, K. and Perring, C. (1990) Carers Medicine, 22, 1295–305. and services. A Review of research.HerMajesty’s 4. Keogh, K.M., White, P., Smith, S.M., et al. (2007) Stationery Office, London. Changing illness perceptions in patients with poorly 20. Robles-Silva, L. (2008) The caregiving trajectory controlled type 2 diabetes, a randomised controlled trial among poor and chronically ill people. Qual Health of a family-based intervention: protocol and pilot study. Res; 18 (3), 358–68. BMC Family Practice, 8, 36. 21. Adamson, J. and Donovan, J. (2005) ‘Normal dis- 5. Department of Health (2001) Diabetes National Service ruption’: South Asian and African/Caribbean relatives Framework (NSF). Department of Health, London. caring for an older family member in the UK. Social 6. Hiscock, J., Legard, R. and Snape, D. (2001) Listening Science and Medicine, 60, 37–48. to Diabetes Service Users: Qualitative Findings for the 22. Godfrey, M. and Townsend, J. (2001) Caring for an Diabetes National Service Framework. Department of Elder with Dementia: the Experience of Asian Care- Health, London. givers and Barriers to the Take-up of Support Services. 7. National Institute of Health and Clinical Excellence Nuffield Institute for Health, Leeds. (NICE). (2007) Type 2 diabetes (update): national clin- 23. Pinquart, M. and Sorensen,¨ S. (2005) Ethnic differences ical guideline for the management in primary and sec- in stressors, resources, and psychological outcomes of ondary care. NICE, London. family caregiving: a meta-analysis. The Gerontologist, 45, 90–106. 8. Roberts, S. (2007) Working Together for better diabetes 24. Katbamna, S., Ahmad, W., Bhakta, P., et al. (2004). Do care. Department of Health, London. they look after their own? Informal support for South 9. Silliman, R.A., Bhatti, S., Khan, A., et al. (1996) The Asian carers. Health and Social Care in the Community, care of older persons with diabetes mellitus: families 12, 398–406. and primary care physicians. Journal of the American 25. Balukonis J, Melkus GD and Chyun D (2008) Grand- Geriatrics Society, 44, 1314–21. parenthood status and health outcomes in midlife 10. Sinclair, A.J. and Bayer, A.J. (2002) Informal carers African American women with type 2 diabetes. Ethn of older adults with diabetes from different ethnic Dis; 18 (2), 141–6. backgrounds. Report to British Diabetic Association 26. Jacobson, S. and Wood, F.G. (2004) Contributions and NHS Executive, West Midlands, UK. of children to the care of adults with diabetes. The 11. Pibernik-Okanovic, M., Rogli, G., Prasek, M., et al. Diabetes Educator, 30, 820–6. (1996). Emotional adjustment and metabolic control in 27. Sayers, S.L., White, T., Zubritsky, C., et al. (2006). newly diagnosed diabetic persons. Diabetes Research Family involvement in the care of healthy medical and Clinical Practice, 34, 99–105. outpatients. Family Practice, 23, 317–24. 12. Toljamo, M. and Hentinen, M. (2001) Adherence to 28. Connell, C.M. (1991). Psychosocial contexts of dia- self-care and social support. Journal of Clinical Nurs- betes and older adulthood: reciprocal effects. The Dia- ing, 10, 618–27. betes Educator, 17, 364–71. 13. Trief, P.M., Wade, M.J., Pine, D., et al. (2003) A 29. Pearlin, L.I., Mullan, J.T., Semple, S.J., et al. (1990). comparison of health-related quality of life of elderly Caregiving and the stress process: an overview of and younger insulin-treated adults with diabetes. Age concepts and their measures. The Gerontologist, 30, and Ageing, 32, 613–18. 583–94. 14. McGarry, J. and Arthur, A. (2001) Informal caring in 30. Archbold, P.F. (1983). Impact of parent-caring on late life: a qualitative study of the experiences of older women. Family Relations, 32, 39–45. carers. Journal of Advanced Nursing, 33, 182–9. 31. Sinclair, A.J., Girling, A.J. and Bayer, A.J. (2000) Cog- 15. Maher, J. and Green, H. (2002) Carers 2000.Officeof nitive dysfunction in older subjects with diabetes mel- National Statistics, London. litus: impact on diabetes self-management and use of 16. Murphy, D.J., Williamson, P.S. and Nease, D.E. (1994) care services. Diabetes Research and Clinical Practice, Supportive family members of diabetic adults. Family 50, 203–12. Practice Research Journal, 14, 323–31. 32. Langa, K.M., Vijan, S., Hayward, R.A., et al. (2002). 17. Hennessy, C.H., John, R. and Anderson, L.A. (1999) Informal caregiving for diabetes and diabetic complica- Diabetes education needs of family members caring for tions among elderly Americans. Journal of Gerontology 32.6 WHAT ARE THE BENEFITS OF CARER INTERVENTION? 451

Series B Psychological Sciences and Social Sciences, 46. Awadalla AW, Ohaeri JU, Tawfiq AM and Al-Awadi 57, S177–86. SA (2006) Subjective quality of life of outpatients with 33. McKinlay, J.B., Crawford, S.L. and Tennstedt, S.L. diabetes: comparison with family caregivers’ impres- (1995) The everyday impacts of providing informal sions and control group. J Natl Med Assoc 98 (5), care to dependent elders and their consequences for 737–45. the care recipients. Journal of Aging and Health, 7, 47. Brod, M. (1998) Quality of life issues in patients 497–528. with diabetes and lower extremity ulcers: patients and 34. Faison, K.J., Faria, S.H. and Frank, D. (1999) Care- caregivers. Quality of Life Research, 7, 365–72. givers of chronically ill elderly: perceived burden. 48. Nabuurs-Franssen, M.H., Huijberts, M.S.P., Nieuwen- Journal of Community Health Nursing, 16, 243–53. huijzen Kruseman, A.C., et al. (2005) Health-related 35. Ekwall, A.K., Sivberg, B. and Hallberg, I.R. (2005) quality of life of diabetic foot ulcers patients and their Loneliness as a predictor of quality of life among older caregivers. Diabetologia, 48, 1906–10. caregivers. Journal of Advanced Nursing, 49, 23–32. 49. Reza, M., Taylor, C.D., Towse, K., et al (2002) Insulin 36. Lee, S., Kawachi, I. and Grodstein, F. (2004) Does improves well-being for selected elderly type 2 diabetic caregiving stress affect cognitive function in older subjects. Diabetes Research and Clinical Practice, 55, women? The Journal of Nervous and Mental Disease, 201–7. 192, 51–7. 50. Holmes, J., Gear, E., Bottomley, J., et al (2003). 37. Jørgensen, H.V., Pedersen-Bjergaard, U., Rasmussen, Do people with type 2 diabetes and their carers lose A.K., et al. (2003). The impact of severe hypoglycemia income? (T2ARDIS-4). Health Policy, 64, 291–6. and impaired awareness of hypoglycemia on relatives 51. Khan, A., Lasker, S.S. and Chowdhury, T.A. (2003) of patients with type 1 diabetes. Diabetes Care, 26, Are spouses of patients with type 2 diabetes at in- 1106–9. creased risk of developing diabetes? Diabetes Care, 26, 38. Gonder-Frederick, L., Cox, D., Kovatchev, B., et 710–12. al. (1997). The psychosocial impact of severe hypo- 52. Hirakawa Y, Kuzuya M, Masuda Y, Enoki H and glycemic episodes on spouses of patients with IDDM. Iguchi A (2008) Influence of diabetes mellitus on Diabetes Care, 20, 1543–6. caregiver burden in home care: a report based on 39. Trief, P.M., Ploutz-Snyder, R., Britton, K.D., et al. the Nagoya Longitudinal Study of the Frail Elderly (2004). The relationship between marital quality and (NLS-FE). Geriatr Gerontol Int; 8 (1), 41–7. adherence to the diabetes care regimen. Annals of 53. Bayer, A. (2004) Telling older patients and their fam- Behavioral Medicine, 27, 148–54. ilies what they want to know. Reviews in Clinical 40. Stodberg,¨ R., Sunvisson, H. and Ahlstrom,¨ G. (2007) Gerontology, 13, 1–4. Lived experience of significant others of persons with diabetes. Journal of Clinical Nursing, 16, 215–22. 54. Levine, C. (1999) The loneliness of the long-term 41. Chappel, N.L. and Reid, R.C. (2002) Burden and caregiver. New England Journal of Medicine, 340, well-being among caregivers: examining the distinc- 1587–90. tion. Gerontologist, 42, 772–80. 55. DiMatteo, M.R. (2004) Social support and patient ad- 42. Fisher, L., Chesla, C.A., Skaff, M.M., et al. herence to medical treatment: a meta-analysis. Health (2002) Depression and anxiety among partners of Psychology, 23, 207–18. European-American and Latino patients with type 2 56. Fisher, L. and Weihs, K.L. (2000) Can addressing fam- diabetes. Diabetes Care, 25, 1564–70. ily relationships improve outcomes in chronic disease? 43. Fisher, L. (2006) Research on the family and chronic Journal of Family Practice. 49, 561–6. disease among adults: Major trends and directions. 57. Karlsen, B., Idsoe, T., Dirdal, I., et al. (2004). Families, Systems, and Health, 24, 373–80. Effects of a group-based counselling programme 44. Li, T.C., Lee, Y.D., Lin, C.C., et al. (2004) Quality of on diabetes-related stress, coping, psychological life of primary caregivers of elderly with cerebrovas- well-being and metabolic control in adults with type 1 cular disease or diabetes hospitalized for acute care: or type 2 diabetes. Patient Education and Counselling, assessment of well-being and functioning using the 53, 299–308. SF-36 health questionnaire. Quality of Life Research, 58. Weihs, K., Fisher, L. and Baird, M. (2002) Families, 13, 1081–8. health, and behavior. Families, Systems, and Health, 45. Awadalla, A.W., Ohaeri, J.U., Al-Awadi, S.A., et al 20, 7–46. (2006) Diabetes mellitus patients’ family caregivers’ 59. Goodall, T.A. and Halford, W.K. (1991) Self- subjective quality of life. Journal of the National Med- management of diabetes mellitus: a critical review. ical Association 98, 727–36. Health Psychology, 10, 1–8. 452 CH 32 SUPPORTING THE FAMILY AND INFORMAL CARERS

60. White, P., Smith, S.M. and O’Dowd, T. (2007). Living of controlled intervention studies. Patient Education with type 2 diabetes: a family perspective. Diabetic and Counselling, 59, 1–12. Medicine, 24, 796–801. 64. Gilden, J.L., Hendryx, M.S., Clar, S., et al. (1992). 61. Harris, M.A. (2006) The family’s involvement Diabetes support groups improve health care of older in diabetes care and the problem of ‘miscarried diabetic patients. Journal of the American Geriatrics helping’. European Endocrine Review, 1–3. Available Society, 40, 147–50. at: www.touchbrieﬁngs.com/pdf/1711/Harris.pdf. 65. Gilliland, S.S., Azen, S.P., Perez, G.E., et al (2002). Accessed 21 December 2007. Strong in body and spirit: lifestyle intervention for 62. Boehm, S., Schlenk, E.A., Funnell, M.M., et al. (1997) Native American adults with diabetes in New Mexico. Predictors of adherence to nutrition recommendations Diabetes Care, 25, 78–83. in people with non-insulin-dependent diabetes mellitus. 66. Wing, R.R., Marcus, M.D., Epstein, L.H., et al. (1991) The Diabetes Educator, 23, 157–65. A family-based approach to the treatment of obese type 63. van Dam, H.A., van der Horst, F.G., Knoops, L., et al. II diabetic patients. Journal of Consulting and Clinical (2005) Social support in diabetes: a systematic review Psychology, 59, 156–62. 33 Diabetes Mellitus Care Models for Older People

I The European Perspective Isabelle Bourdel-Marchasson Geriatric Department, Hopitalˆ Xavier Arnozan, CHU of Bordeaux, Pessac cedex, UMR 5536 CNRS/Universite´ Victor Segalen Bordeaux 2, Bordeaux, France II The United States Perspective John E. Morley GRECC, VA Medical Center and Division of Geriatric Medicine, Saint Louis University School of Medicine, St Louis, MO, USA I The European Perspective

expectancy, this will inevitably result in a higher Key messages prevalence of diabetes, particularly among older peo- • The increasing numbers of older subjects with di- ple. At the beginning of the twenty-ﬁrst century, ap- abetes will lead to a further public health burden proximately 50% of all French or British diabetic sub- associated with corresponding socioeconomic jects receiving treatment were aged >65 years, and burden. 25% were aged >75 years [1, 2]. Similar proportions • Relatively little research data are available in the among the diabetes population were identiﬁed in the area of quality of life and health status in older Netherlands during the same time period, with some people with diabetes. 50% of people with type 2 diabetes aged >70 years [3]. • Enhancing diabetes care for older people re- This change in the proportion of people with diabetes quires well-designed and focused diabetes au- has numerous implications, most notably the problems dits and the evaluation of differing models of of associated costs and the quality of care. In France, care. in 2004, the mean annual health insurance reimbursement for diabetes was ¤5910 per person, this value having increased slightly during the past 10 years. Yet, 33.1 Introduction the growth in the proportion of diabetic subjects in any age group was estimated at 6% per year, leading Today, European countries are facing an increasing to expanding health care costs [4]. Direct medical ex- rate of obesity and sedentary lifestyle among their penses were also found to be higher in subjects aged populations. In conjunction with an increasing life >70 years than in younger subjects in the Netherlands

(¤2080 per year in those aged >70 years versus ¤1040 study described above [7], the proportion of diabetic in those aged <50 years) [5]. patients with a HbA1c level >8% was similar in The efficiency of current diabetes care models in the both known or unknown diabetes (respectively 19.6% prevention of complications requires assessment espe- and 16.1%). The criteria for screening include the cially in the elderly. Indeed, the lack of evidence in presence of symptoms of hyperglycaemia, or the the field of geriatric diabetes – and particularly in those presence of conditions that are frequently associated aged >75 years – makes the provision of clinical guid- with diabetes. Unfortunately, however, diabetes ance a difficult proposition. Assuming that one-quarter is frequently asymptomatic or the symptoms are of the diabetic population may be aged >75 years in non-specific, such as fatigue, weight loss and mood Europe, major efforts towards efficiency analysis are changes. As a result, diabetes screening should also now required in different European countries. be instigated when other geriatric problems emerge such as falls, incontinence, mental alterations and pain [6]. According to the European guidelines mentioned 33.2 Current clinical guidelines above, the frequency of screening for diabetes in in Europe those subjects with one or more risk factors should be at two-yearly intervals for those aged 65–74 years, When the European Diabetes Working Party for Older and annually for those aged ≥75 years. The Desktop People (2001–2004) developed evidence-based clini- Guide to Type 2 Diabetes Mellitus, as approved by cal guidelines dedicated to the elderly, they were in the IDF (European Region), has recommended that fact designed to complement the existing guidelines of an oral glucose tolerance test (OGTT) be conducted the International Diabetes Federation (IDF, European particularly in the elderly, when the fasting plasma Region), and the European Association for the Study glucose (FPG) is normal but the suspicion is high, or of Diabetes (EASD) [6]. The founding statement was there is a high risk of diabetes. the promotion of overall well-being and normal life expectancy, and in particular the maintenance of an 33.2.2 Healthy ageing, frailty optimal level of physical and cognitive function. The and pathological ageing right of older diabetic patients to have access, with con- Whilst much ethical discussion has arisen – and confidence, to skilled care was also proposed. However, tinues to arise – regarding the problem of prevention in the position of prevention of diabetes – whether sec- the frail elderly, it I important first to clarify the target ondary or tertiary – must be clarified for each elderly population for intensive multifactorial and expensive patient in a similar manner as for younger patients. prevention programmes. Furthermore, such intervention might be problematic in frail elders, in which 33.2.1 The screening of diabetes frailty represents the state between full autonomy or in the elderly successful ageing and irreversible functional depen- The early detection of diabetes in the elderly dency, or pathological ageing. Yet, while this defini- could improve the efficiency of prevention of tion is imprecise the potential clinical criteria are both diabetes-induced disabilities. Undiagnosed diabetes numerous and interrelated [9]. Indeed, diabetes could could involve one-third of elderly diabetic people, itself be considered as a frailty factor due to the mul- although this proportion seems to vary widely ac- tifaceted association between diabetes and disability cording to the country of origin. For example, among [10]. Frailty is therefore not necessarily associated with British institutions one study estimated that 16.9% a high mortality risk since, beyond frailty guidelines of residents had undiagnosed diabetes compared for the care of subjects with irreversible functional de- to 12.1% with known diabetes [7]. In contrast, pendency are lacking. Indeed, in this later category, among a French community-based population, the prevention is likely to be of a tertiary nature, with em- proportion of undiagnosed diabetes was only 1.4%, phasis placed on daily living and the quality of life. compared to 8.2% with known diabetes [8]. Moreover, 33.2.3 Construction of targets undiagnosed diabetes does not necessarily mean that for diabetes control the condition is ‘mild’, as both microvascular and macrovascular complications are invariably present In European countries, few diabetes care guidelines at the time of diagnosis. In the French community have addressed the issue of control in elderly diabetic 33.2 CURRENT CLINICAL GUIDELINES IN EUROPE 455 patients. Targets for control and preventive actions also be identified [6]. Caregivers, whether professional could be derived from the categorization according or familial, should receive diabetic education, and for to frailty criteria. According to the European Diabetes this purpose a number of dietetic guidelines for care- Working Party for Older People, in older patients with givers have been derived. In France, for example, a single system involvement (i.e. free of other major co- mini-guide for caregivers of subjects with different morbidities), a target HbA1c level of 6.5–7.5% should health problems (including diabetes) is currently avail- be aimed for [6]. However, the precise target agreed able (at no cost) from the AFSSA (the French agency will depend on any existing cardiovascular risk, the for the health security of food). presence of microvascular complications, and the ability of the individual to self-manage. For those subjects 33.2.5 The rationale for Comprehensive who are dependent, suffering from multisystem dis- Geriatric Assessment (CGE) in frail ease, living in care home residency (including those elderly diabetic subjects with dementia), and those patients in whom the hypoglycaemia risk is high and symptom control and There are several arguments for performing an ex- avoidance of metabolic decompensation paramount, tended CGE in elderly diabetic patients. In general, the the target HbA1c range should be >7.5% to 8.5% CGE is recommended in frail elderly, with diabetes be- [6]. Thus, targets for control are highly dependent ing considered a frailty factor, much like cancer [9]. on a benefit:risk ratio assessment, which itself may A CGE offers procedures to address the needs of frail be related to iatrogenic problems and difficulties in subjects, with the aim of increasing their quality of life self-management. The most recent French guidelines and preventing any worsening of their functional de- for diabetes treatment have recommended the identi- pendency. The CGE targets multidisciplinary domains fication of three types of elderly subject: (i) in the and approaches, including medical, social and psycho- case of end of life and a necessity for comfort, in- logical fields. Indeed, diabetic subjects are prone to sulin treatment should be used; (ii) in the case of have social difficulties that interfere with both dia- poly-pathology, target of 8% should be preferred; and betes risk and diabetes care. The CGE also provides (iii) in the case of patients with a single system involve- a good opportunity for controlling polypharmacy and ment, the extrapolation of data obtained in younger preventing adverse drug reactions and interactions. On subjects seems reasonable, with a recommendation of the other hand, it might be considered a waste of time starting treatment with reduced dosages, the use of and money to perform the CGE in either vigorous peo- Cockcroft formulae and respect of contraindications ple or in those with irreversible functional and mental [11]. The clinical guide for the elderly diabetic patient dependency. Prior to the CGE, the identification of frail care among a French-speaking group for study of dia- subjects is therefore necessary, and consequently frailty betes in the elderly, has incorporated the same strategy criteria have been proposed (Table 33.1) [6, 9, 12]. for control target determination as did the European In clinical practice, the CGE could take on different guide in 2004 [12]. frameworks related to the level of complexity of the patient’s situation. The minimum staff requirement to 33.2.4 Therapeutic education perform a CGE includes a physician, a nurse and a social worker, although the team is greatly and advan- Access to adequate support service – and particularly tageously enriched by the presence of a psychologist, to therapeutic education – are considered the rights of a rehabilitation therapist, a dietician and a dental sur- elderly diabetic patients. It has been shown that an geon. Access to basic biochemistry and haematological 8-week education programme utilizing telemedicine analyses and X-ray examinations is also required. could be applied to elderly subjects (mean ± SD age 73.3 ± 6.9 years), and could also increase the sub- 33.2.6 Specificity of institutions jects’ quality of life [13]. The provision of diabetes education could be limited due to sensorial, cognitive Today, an increasing number of older people are liv- and mobility impairments. Moreover, the expectations ing in geriatric institutions. A recent meta-analysis has and beliefs of older people could also be altered, thus shown that main risk factors for admission were im- complicating the communication process. Therefore, it pairment of at least two basic activities of daily living, is recommended that, besides educational needs, any cognitive impairment, social isolation and low income, significant barrier to the receipt of education should and prior hospitalization or nursing home residence 456 CH 33 DIABETES MELLITUS CARE MODELS FOR OLDER PEOPLE

Table 33.1 The criteria for frailty screening [6, 9, 12]. agent. Diabetes-related symptoms, such as hyperosmolar conditions, could be prevented. In 1999, the British Category Criteria Diabetic Association (now known as Diabetes UK) Age Continuous effect after age 65 proposed the introduction of speciﬁc guidelines for years care homes [15]. Morbidity Stroke These recommendations state that each resident Chronic and disabling disease should be screened for the presence of diabetes, at Cancer admission to the home, and then annually. An individ- Absence of terminal illness ualized care plan should subsequently be implemented Diabetes with a blood glucose control target, a dietary plan, a Polypharmacy and complex drug suggested body weight and a nursing plan. Any dia- regimens betic resident should then undergo an annual review, Functional Need for help with shopping, impairment cooking and housekeeping and have access to the following services: ophthalmol- Mobility impairment ogy, diabetes foot clinic, dietetic and diabetes specialist Slow gait speed team (nurse and physician). Sensorial impairment Geriatric Delirium 33.2.7 Physician and staff training syndromes to diabetes care Under-nutrition Falls There are moves in hand to increase the level of Incontinence diabetes education and training among health care staff. Pressure ulcer It is thus recommended to enhance initial teaching Increased susceptibility to adverse and postgraduate training to address these new and drug events increasing needs of skilled prevention and care. The Mental health Depression initiation of insulin therapy necessitates a large degree Dementia from mild to moderate of competency. In the elderly (notably among the frail (severe excluded) elderly) and in institutions, the rate of insulin therapy Nursing Restraints use is particularly high and recommended, which again Bed-ridden emphasizes the importance of training in this area. It General Poor subjective health has been suggested that, in each geriatric institution, Social Socioeconomic problems diabetes education and training courses for staff should Familial difﬁculties be available [6].

[14]. In a geriatric setting, the prevalence of diabetes will most likely be higher than in the community, 33.3 Care organization due largely to the role of diabetes in the disablement process in elderly. Thus, specific guidelines for care Today, in the field of diabetes care, professional net- home residency have been proposed [6]. works are now either currently in operation or are The baseline concerns were a lack in specialist under development. In France, the national associa- follow-up, an inadequate dietary care, a lack of in- tion of regional networks for diabetes care (ANCRED) dividualized diabetes care plans, and a lack of any includes professionals and patients with the aim of pro- educational and training programmes for the care home viding to patients the best care they need, where they staff. Until now, no significant intervention study of live. Such networks were constructed to help the appli- any type has been conducted to assess the benefit of cation of current guidelines. Despite this, the specific metabolic control and educational strategy in institu- needs of vulnerable elders are addressed very little, tionalized elderly diabetic patients. However, the Eu- which again echoes the minimal portion of the guide- ropean guidelines recommended an early detection of lines that is dedicated to the elderly. diabetes in elderly [6], where the field is not primary Thus, the European guidelines [6] have proposed a prevention but rather the diagnosis of diabetes at an framework for a management system of diabetes in early stage. This could lead to early dietary intervention the elderly, with the aim of implementing guidelines and perhaps then delay the use of an hypoglycaemic in clinical practice. Such management systems should 33.4 CALL FOR AUDIT OF QUALITY OF CARE 457

Audit

Continuing professional development

Geriatrician for Diabetes care team GP CGE Standards and protocol of care Patient specific care plan Annual review recall system Needs-based Anonymous data education model patient

Eye screening Access to care liaison service Structured record High quality system laboratory service Services directly accessible to older even with disabilities Adapted education

Figure 33.1 Recommendations for a diabetes management system dedicated to elderly diabetic subjects. GP = General practitioner; CGE = Comprehensive Gerontological Evaluation. function in networks and provide training and measures Table 33.2 Audit of Quality of care for older diabetic through planned audits of application and effects. In people. the elderly, the criteria for quality of care should include the rates of disability and institutionalization Target point Rationale (Figure 33.1). Assessment of criteria for Acquisition of baseline diabetes care according data on quality of care to current guidelines 33.4 Call for audit of quality of care for the young Description of subjects Designing prevention The European Diabetes Working Party for Older Peo- according to frailty procedures according ple in 2004 called for audits of care in elderly diabetic criteria and diabetes to subjects’ health populations. The level of evidence for guidelines in complications status all along the this population is particularly low, and observational audit process studies are required to assess the quality of care with reference to current guidelines, and to follow the ef- Description of actual care Improving quality of care fects on such quality of care. Although quality of care networks around older in continuous subjects management based on audits in the general population with reference to cur- available resources rent guidelines have been widely performed in Europe, very little attention has been paid to the oldest subjects. Assessment of functional Assessment of Some target points for audits in diabetes care for older outcome performance at patient people are listed in Table 33.2. level ENTRED 2001 was an observational study of a Including professionals of Optimizing feedback representative sample of 10 000 French diabetic pa- care in the audit from/to the tients constructed on the basis of anti-diabetic drug process professionals 458 CH 33 DIABETES MELLITUS CARE MODELS FOR OLDER PEOPLE reimbursement from the National Health Service [16]. true for any chronic condition [20]. In a French audit of As mentioned above, one-quarter of these patients were a group of geriatric care homes, various procedures of aged >75 years. Subjects and their general practition- medical follow-up were performed as recommended in ers were asked to complete a questionnaire and, for only 25–50% of residents. Among those items, HbA1c each patient, reimbursements of medical consultations was tested at least three times each year in 25% of res- and biological analysis were recorded. The response idents, while another 25% had no HbA1c measurement rate to self-questionnaires was low in older subjects at all, 40% had no weight measurement, and 47% had compared to younger, with a linear shape decrease no blood pressure recorded within the past 18 months. from 50% in those aged <65 years to 25% in those The level of HbA1c was poorly controlled (>8.5%) aged >85 years. This audit showed a low referral to in 20% of the people, which was close to the esti- specialists in the older subjects, and also a lower level mate of 17% found in a British study [21]. The other of follow-up. A lack of any HbA1c measurement was items included blood pressure and weight monitoring, found in 11% of subjects aged >75 years, in 6% in while annual eye, foot, cardiovascular, renal and den- those aged 65–74 years, and in 7% in those aged <64 tal examinations were also performed to an insufficient years. A global analysis showed that the care of sub- degree. jects aged >65 years displayed a tertiary prevention pattern. An audit of quality of care performed in diabetes 33.5 Perspectives centres in Italy showed that the application of current guidelines was different according to centres, and that The future outlook suggests that there is a need to this was the most important cause of variation [17]. In increase the evidence base in geriatric diabetes, and Italy, the aim of the IMPLEMEG study was to assess also to ensure that the guidelines provide sound and the efficiency of different strategies for the implemen- robust assistance in delivering high-quality diabetes tation of diabetes care guidelines [18]. However, inves- care.Thefirststepinthisdirectionwouldseemto tigations of the effect of age and comorbidity on care be the establishment of pragmatic guidelines that are choices were not planned. This issue seems an inter- specific to elderly subjects, but with special attention esting point to explore, and it has indeed been shown being paid to frail or dependent subjects. In addition, that age and comorbidity reduced the rate of intensifi- these guidelines should be periodically reassessed on cation of therapy [19]. In fact, in this latter study no the basis of both cohort follow-up studies and on reference to the CGE was made, and the quality of investigations of both feasibility and cost-effectiveness. care audit was based on guidelines for the general dia- A minimum data set has been recently established via betic population in a secondary prevention perspective, Gerontonet, a European network for the study of health which was unlikely to have been adapted to all of the care among older people. This comprised a short CGE oldest patients. which was particularly fitted to diabetic subjects [22]. The quality of care provided to elderly people with In old patients, end points such as quality of life and diabetes living in long-term care facilities is generally disability-free expectancy should be preferred, rather insufficient, and worse than that provided to people liv- than survival. ing in their own home, although this may of course be 33.7 SPECIFIC ISSUES IN OLDER DIABETICS 459 II The United States Perspective

Table 33.4 Complications in older persons with dia- Key messages betes mellitus. • The prevalence of diabetes mellitus in older Complication Prevalence (per 1000) people within the United States has doubled over the past 25 years. Aged 65–74 Aged 75+ • Both, metformin and thiazolidinediones, are years years increasingly used in the United States to treat Mobility impairment 662 783 older people with type 2 diabetes. Cardiovascular disease 480 580 Visual impairment 226 271 Peripheral arterial disease 8 12 Ulcers 8 10 33.6 Introduction End-stage renal disease 4 4 Neuropathy 4 4 In the United States, 12.2 million persons aged >60 years have diabetes mellitus and thus, over half of all diabetics are aged >60 years. Currently, over half a after 5 years in older diabetics compared to 15 years million new cases of diabetes mellitus are diagnosed in younger diabetics. Cardiovascular disease is present each year in persons aged over 60 years, with the mean in 58% of diabetics aged >75 years, while mobility age of onset of the condition being 51.9 years. The is impaired in 78.3% and visual impairment occurs in prevalence of diabetes mellitus in persons aged >65 27.1% of diabetics aged >65 years. Amputations occur years doubled between 1980 and 2005 (Table 33.3). commonly in diabetics aged ≥65 years; each year, 3.4 The rate of increase in diabetes mellitus has been subjects in 1000 have above-knee amputations, 3.3 in greater than the rate of increase in obesity. On an ethnic 1000 below-knee amputations, and one in 1000 a foot basis, diabetes is almost twice as common in Blacks amputation. than Whites, and 60% more common in Hispanics, In two national studies examining the prevalence of while older males are more likely to be diabetic than diabetes mellitus in nursing homes [24, 25], one study are females. Among persons with diabetes mellitus reported a value of 24.6%, and the other 26.4%, which aged 65–79 years, the average duration of the condition was similar to the value of 23.8% seen in the gen- is 9.6 years. Moreover, diabetes is the seventh leading eral population aged >60 years. The present author’s cause of death in persons aged >65 years. unpublished data have suggested that the prevalence Diabetic complications are more common in older of diabetes mellitus may be closer to 30% in nursing persons (Table 33.4). For example, Rosenthal et al. homes when diagnoses are carefully made, instead of [23] found that nephropathy and retinopathy occurred relying on the physicians’ diagnoses. This is in keeping with the prevalence of undiagnosed diabetes mellitus in Table 33.3 The changing prevalence of the community. Typically, those people with diabetes diabetes mellitus in the United States, mellitus who are resident in nursing homes receive between 1980 and 2005. more medications (mean 10.3) than those non-resident Year Prevalence (%) (mean 8.4).

Aged 65–74 Aged 75+ years years 33.7 Specific issues in older 1980 9.1 8.9 diabetics 1985 10.2 10.0 1990 9.9 8.6 Cognitive problems are extremely common in 1995 11.1 10.6 older diabetics [26], and clearly this is, in part, 2000 15.4 13.0 associated with an increase in vascular dementia 2005 18.5 15.6 secondary to the increase in atherosclerosis. In 460 CH 33 DIABETES MELLITUS CARE MODELS FOR OLDER PEOPLE addition, hyperinsulinaemia is associated with an Table 33.5 Special issues in the care increase in Alzheimer’s disease, possibly because the of the older diabetic patient. insulin-degrading enzyme also degrades amyloid-beta • Cognitive impairment peptide in the brain. Both, hyperglycaemia and hypo- • glycaemia are associated with memory disturbances; Vascular dementia indeed, it was recently shown that hypertriglyceri- • Alzheimer’s disease daemia causes a direct decrease in memory capability • Hypoglycaemia [27]. Cognitive disturbances in persons with diabetes • Hyperglycaemia lead to compliance issues, and for these reasons it is • suggested that all older diabetics have their cognitive Hypertriglyceridaemia function regularly checked, using a tool such as the • Depression Saint Louis University Mental Status Test [28]. • Falls Depression is more common among older diabetics • Fractures than among other older persons, and is largely associated with an increase in hospitalizations and mortality • Sarcopenia [23]. Depression is also a cause of worse compliance • Disability in older persons. • Pressure ulcers Falls are a major cause of injuries in older persons, • Nocturia with injurious falls being more common among older diabetics than among others living in the community • Sleep disturbances [29]. Among people with diabetes, falls tend to be more • Incontinence common when the HbA1c level is <7% [30]; the causes • Tuberculosis of such falls include hypoglycaemia, orthostasis, syn- • cope, medications, peripheral neuropathy, myopathies, Recurrent urinary tract infection balance problems, foot abnormalities, impaired vision • Hypogonadism and impaired executive function leading to problems with dual tasking (a failure to be able to ‘walk and in cytokine levels, and this leads to an accelerated mus- talk’). Therapy with statins may be an important cause cle loss [37]. Insulin resistance leads to an increased fat of falls in older diabetics. accumulation in muscle and decreased muscle power, Pressure ulcers are more common and heal more while atherosclerosis leads to peripheral vascular dis- slowly in older persons with diabetes mellitus [31, ease and muscle anoxia. Sarcopenia results in increased 32], possibly due to zinc deficiency, to tissue glycation functional impairment in older diabetics [38]. and vascular impairment. Zinc deficiency is common High glucose levels represent an excellent medium in diabetics as people with diabetes mellitus have an for the growth of bacteria and fungi; consequently, impaired absorption of zinc and consequent hyperzin- older diabetics are highly prone to recurrent tuber- curia [33]. culosis, to suffer fungal infections and have recurrent People with type 2 diabetes mellitus tend to have urinary tract infections. an increase in numbers of fractures, despite having Some of the special issues that must be considered an increase in their bone mineral density (BMD) [34]. in the older diabetic patient are listed in Table 33.5. Thiazolidinediones tend to reduce BMD, with the result being a greater incidence of fracture [35]. Hyperglycaemia results in a hyperosmolar diuresis 33.8 Management of diabetes that is associated with nocturia and sleep disturbances, as well as incontinence. In many cases, however, good The treatment of older diabetics tends to be relatively diabetic control can decrease nocturia and inconti- aggressive in the United States where, in nursing nence. homes, many patients will have serum HbA1c levels Diabetics tend to have an accelerated loss of muscle <7% [39]. Among the older diabetic community there mass (sarcopenia) that occurs with ageing. This may be is also a greater use of medications in persons aged due to a greater prevalence of hypogonadism in older >65 years (88%) compared to those aged 18–44 years diabetic males [36]. Diabetics may also show increases (71.7%). 33.9 CONCLUSIONS 461 In older diabetics, weight loss has been shown to loss of bone have limited its use in this population. be associated with increased mortality [40], although The oedema associated with rosiglitazone has also in various institutions it has been shown that neither markedly limited its use, while both exenatide and diabetic diets nor the restriction of sweets improves pramilentide – both of which cause weight loss – are diabetic control [41, 42]. rarely used in older persons because of their potential In the Diabetes Prevention Program, a lifestyle mod- interaction with the anorexia of aging [45]. ification with a focus on exercise was most effective The use of sitagliptin, a DPP-IV inhibitor, which in preventing diabetes in the 60–85-year-old group increases the incretin, glucagon-like peptide, is rapidly (3.3 per 100 persons-years) compared to that in the increasing. Interestingly, there has also been some 45–59-year olds (4.9 per 100 persons-years) and the use – albeit minimal – of the two alpha-1-glucosidase 25–44-year olds (6.3 per 100 persons-years) [43]. Met- inhibitors which also increase glucagon-like peptide formin, on the other hand, was less effective with [46, 47]. advancing age [43]. Between 1990 and 2001 there was a 2.9-fold increase in the use of oral anti-diabetic drugs in the 33.9 Conclusions United States, due mainly to the introduction of metformin in 1995, followed by the thiazolidinediones. By Overall in the United States, diabetes mellitus is rela- 2007, the thiazolidinediones accounted for 6.3 million tively aggressively treated, especially in institutional- prescriptions compared to 1.4 million for glyburide. ized older persons. The National Health Objectives for A subsequent meta-analysis suggested that rosigliti- 2010 require: (i) a dilated eye examination in 75% zone was associated with increased cardiovascular dis- of cases; (ii) a foot examination in 75%; (iii) the ease, and this led to a rapid decrease in the drug’s measurement of plasma glucose daily in 60%; and prescriptions. Although, in 2007 the long-acting in- (iv) an annual assessment of serum HbA1c level in sulin gla-insulin (lantus) was the second best selling 50% of cases. Among those persons aged >65 years anti-diabetic drug in the United States, during the same by the year 2000, almost 72% had undergone a di- year the total sales of the first of the dipeptidyl pepti- lated eye examination and 55% a foot examination. dase IV (DPP-IV) inhibitors reached US$ 471 million. The self-monitoring of blood glucose in persons aged The anti-diabetic drugs most commonly used in the 65–74 years was 43%, while HbA1c levels were mon- United States are listed in Table 33.6. Long-acting itored in 15.9%. Moreover, the proportions were even insulin, which is now commonly administered as a less than in those subjects aged >75 years. Based on basal insulin in older persons, can be used with or the NHANES data, the mean serum HbA1c level in per- without other oral anti-diabetic drugs or shorter-acting sons aged >65 years with diabetes was reduced from insulins. Metformin is generally not used in the United 7.49% (i.e. <7% in 36.2% of cases) in 1999–2000 to States in persons aged >80 years, as the majority 6.72% (<7% in 68.3% of cases) in 2003–2004 [48]. of such patients will have some degree of impaired The management of hypertension and lipids was also renal function. As calculated by the Cockcroft–Gault improved. Despite these findings, the results of some equation, metformin also causes anorexia and weight recent studies have proposed that the excessive con- loss [44]. trol of diabetes mellitus might in fact be deleterious, Although pioglitazone is still commonly used in which suggests that the care of older diabetics may be ‘old-old’ patients, its potential effects to enhance the approaching optimal levels in the United States.

Table 33.6 Drugs used to treat diabetes mellitus in older persons in the United States. Insulins Sulphonylureas Metformin α-Glucosidase Thiazolidenediones DPP-IV inhibitors inhibitors Target Muscle/fat Beta cell Liver Gut/beta cell Muscle/fat Beta cell HbA1c >2% 1–2% 1–2% 0.5–1% 0.5–2% 0.5–1% Effect on fasting Positive Positive Positive Moderate Positive Moderate Postprandial effect Positive Moderate Moderate Positive Moderate Positive Weight change Increase Increase Decrease Decrease Increase No change Cost (US$) 30–450 10–15 30–60 40–80 80–100 200 462 CH 33 DIABETES MELLITUS CARE MODELS FOR OLDER PEOPLE Acknowledgments treatment of type 2 diabetes. Recommendations for good practice]. Diabetes Metab. 2007; 33 (1Pt 2): The author acknowledges that the statistics on di- 1S7–25. 12. French-speaking group for study of diabetes in the abetes in the United States were abstracted from elderly. (2008) Guide pour la prise en charge du www.cdc.gov in June, 2008. diabetique´ agˆ e´ [Clinical guide for elderly diabetic patient care]. Medecine´ des Maladies Metaboliques´ 2008; 2(Horsserie):´ 1–53. References 13. Chan WM, Woo J, Hui E, Lau WW, Lai JC and Lee D. A community model for care of elderly people with 1. Diabetes in the UK (2004) Diabetes, UK, London. diabetes via telemedicine. Appl Nurs Res. 2005; 18 (2): 2. Nevanen´ S, Tambekou J, Fosse S, Simon D, Weill A, 77–81. Varroud-Vial M, et al. Caracteristiques´ et etat´ de sante´ 14. Gaugler JE, Duval S, Anderson KA and Kane RL. des personnes diabetiques´ agˆ ees´ et leur prise en charge Predicting nursing home admission in the U.S: a medicale,´ etude´ Entred 2001. Bulletin Epidemiologique meta-analysis. BMC Geriatrics. 2007; 7: 13. Hebdomadaire 2005; 12–13: 51–2. 15. British Diabetic Association. (1999) Guidelines of 3. Ubink-Veltmaat LJ, Bilo HJ, Groenier KH, Houweling Practice for Residents with Diabetes in Care Homes. ST, Rischen RO and Meyboom-de Jong B. Prevalence, Available from: www.diabetes.org.uk/Documents/ incidence and mortality of type 2 diabetes mellitus re- Reports/guideline residents.pdf. (accessed 1 November visited: a prospective population-based study in The 2007). Netherlands (ZODIAC-1). European Journal of Epi- 16. Institut National de veille sanitaire. (2005) ENTRED demiology 2003; 18 (8): 793–800. study. Representative sample of French diabetic 4. Vallier N, Weill A, Salanave B, Bourrel R, Cayla M, people. Available from: http://www.invs.sante.fr/ Suarez C, et al. Cost of thirty long-term diseases for recherche/index2.asp?txtQuery=Entred (accessed 1 beneﬁciaries of the French general health insurance November 2007). scheme in 2004. Pratiques et Organisation des Soins 17. Petrelli A, Saitto C, Arca M and Perucci CA. Diabetes 2006; 37 (4): 267–83. management by hospital-afﬁliated diabetes centres in 5. Redekop WK, Koopmanschap MA, Rutten GE, Wolf- fenbuttel BH, Stolk RP and Niessen LW. Resource Lazio, Italy. European Journal of Public Health 2004; consumption and costs in Dutch patients with type 2 14 (2): 120–2. diabetes mellitus. Results from 29 general practices. 18. Perria C. Strategies for the introduction and imple- Diabet Med. 2002; 19 (3): 246–53. mentation of a guideline for the treatment of type 2 6. European Diabetes Working Party for Older People diabetics by general practitioners (GPs) of the Lazio re- (2004) Clinical Guidelines for Type 2 Diabetes Mel- gion of Italy (IMPLEMEG study): protocol for a cluster litus. Available from: http://www.eugms.org/index.php randomised controlled trial [ISRCTN80116232]. BMC (accessed 1 November 2007). Health Services Research 2004; 4 (1): 13. 7. Sinclair AJ, Gadsby R, Penfold S, Croxson SC and 19. Chaudhury SI, Berlowitz DR and Concato J. Do age Bayer AJ. Prevalence of diabetes in care home resi- and comorbidity affect intensity of pharmacological dents. Diabetes Care 2001; 24 (6): 1066–8. therapy for poorly controlled diabetes mellitus? Jour- 8. Bourdel-Marchasson I, Helmer C, Barberger-Gateau P, nal of the American Geriatrics Society 2005; 53 (7): Peuchant E, Fevrier B, Ritchie K, et al. Characteristics 1214–16. of undiagnosed diabetes in community-dwelling French 20. Fahey T, Montgomery AA, Barnes J and Protheroe elderly: the 3C study. Diabetes Research and Clinical J. Quality of care for elderly residents in nursing Practice 2007; 76 (2): 257–64. homes and elderly people living at home: controlled 9. Bourdel-Marchasson I and Berrut G. Caring the elderly observational study. Br Med J (Clinical Research Ed.) diabetic patient with respect to concepts of successful 2003; 326 (7389): 580. aging and frailty. Diabetes Metab. 2005; 31 (Spec. No. 21. Taylor CD and Hendra TJ. The prevalence of diabetes 2): 13–19. mellitus and quality of diabetic care in residential and 10. Bourdel-Marchasson I, Helmer C, Fagot-Campagna A, nursing homes. A postal survey. Age and Ageing 2000; Dehail P and Joseph PA. Disability and quality of life 29 (5): 447–50. in elderly people with diabetes. Diabetes Metab. 2007; 22. Sinclair AJ. Towards a minimum data set for interven- 33 (Suppl. 1): S66–74. tion studies in type 2 diabetes in older people. Journal 11. Traitement medicamenteux´ du diabete` de type of Nutrition, Health and Aging 2007; 11 (3): 289–93; 2. Recommandation de bonne pratique [Medical discussion 92–3. 33.9 CONCLUSIONS 463

23. Rosenthal MJ, Fajardo M, Gilmore S, et al. Hospi- associated with type 2 diabetes mellitus: a common talization and mortality of diabetes in older adults. defect? Metab Syndr Relat Disord. 2003, 1, A 3-year prospective study. Diabetes Care 1998; 21: 75–80. 231–5. 37. Kim MJ, Rolland Y, Cepeda O, et al. Diabetes mellitus 24. Resnick HE, Heineman J, Stone R and Shorr RI. (2008) in older men. Aging Male 2006, 9, 139–47. Diabetes in U.S. Nursing Homes, 2004. Diabetes Care, 38. Morley JE, Kim MJ, Haren MT, et al. Frailty and the 31, 287–8. aging male. Aging Male 2005, 8, 135–40. 25. Travis SS, Buchanan RJ, Wang S and Kim M. Analyses 39. Joseph J, Koka M and Aronow WS. Prevalence of a of nursing home residents with diabetes at admission. hemoglobin A1c less than 7.0%, of a blood pressure J Am Med Dir Assoc. 2004, 5, 320–7. less than 130/80mmHg, and of a serum low-density 26. Biessels GJ, Deary IM and Ryan CM. Cognition and lipoprotein cholesterol less than 100mg/dL in older diabetes: a lifespan perspective. Lancet Neurol. 2008, patients with diabetes mellitus in an academic nursing 7, 184–90. home. J Am Med Dir Assoc. 2008, 9, 51–4. 27. Farr SA, Yamada KA, Butterfield DA, et al.Obe- 40. Wedick NM, Barrett-Connor E, Knoke JD and Wingard sity and hypertriglyceridemia produce cognitive im- DL. The relationship between weight loss and all-cause pairment. Endocrinology 2008, 149, 2628–36. mortality in older men and women with and without 28. Tariq SH, Tumosa N, Chibnall JT, et al. Comparison of diabetes mellitus: the Rancho Bernardo study. J Am the Saint Louis University mental status examination Geriatr Soc. 2002, 50, 1810–15. and the mini-mental state examination for detecting 41. Tariq SH, Karcic E, Thomas DR, et al. The use of a dementia and mild neurocognitive disorder – a no-concentrated-sweets diet in the management of type pilot study. Am J Geriatr Psychiatr. 2006, 14, 2 diabetes in the nursing homes. J Am Diet Assoc. 900–10. 2001, 101, 1463–6. 29. Miller DK, Lui LY, Perry HM, III, et al. Reported 42. Coulston AM, Mandelbaum D and Reaven GM. Di- and measured physical functioning in older inner-city etary management of nursing home residents with diabetic African Americans. J Gerontol A Biol Sci Med non-insulin-dependent diabetes mellitus. Am J Clin Sci. 1999, 54, M230–6. Nutr. 1990, 51, 67–71. 30. Nelson JM, Dufraux K and Cook PF. The relationship 43. Drandall J, Schade D, et al. and the Diabetes Prevention between glycemic control and falls in older adults. J Program Research Group. The influence of age on Am Geriatr Soc. 2007, 55, 2041–4. the effects of lifestyle modification and metformin in 31. Brandeis GH, Ooi WL, Hossain M, et al. A longitudinal prevention of diabetes. J Gerontol A Biol Sci Med Sci., study of risk factors associated with the formation of 2006, 61, 1075–81. pressure ulcers in nursing homes. J Am Geriatr Soc. 44. Lee A and Morley JE. Metformin decreases food con- 1994, 42, 388–93. sumption and induces weight loss in subjects with obe- 32. Berlowitz DR, Brandeis GH, Morris JN, et al. Deriving sity with type II non-insulin-dependent diabetes. Obes a risk-adjustment model for pressure ulcer development Res. 1998, 6, 47–53. using the Minimum Data Set. J Am Geriatr Soc. 2001, 45. Rolland Y, Kim MJ, Gammack JK, et al.Officeman- 49, 866–71. agement of weight loss in older persons. Am J Med. 33. Neiwoehner CB, Allen JI, Boosalis M, et al.Roleof 2006, 119, 1019–26. zinc supplementation in type II diabetes mellitus. Am 46. Lee A, Patrick P, Wishart J, et al. The effects of miglitol J Med. 1986, 81, 63–8. on glucagon-like peptide-1 secretion and appetite sen- 34. Dobnig H, Piswanger-Solkner JC, Roth M, et al. Type sations in obese type 2 diabetics. Diabetes Obes Metab. 2 diabetes mellitus in nursing home patients: effects 2002, 4, 329–35. on bone turnover, bone mass, and fracture risk. J Clin 47. DeLeon MJ, Chandurkar V, Albert SG and Mooradian Endocrniol Metab. 2006, 91, 3355–63. AD. Glucagon-like peptide-1 response to acarbose in 35. Meier C, Kraenzlin ME, Bodmer M, et al.Useof elderly type 2 diabetic subjects. Diabetes Res Clin thiazolidinediones and fracture risk. Arch Intern Med. Pract. 2002, 56, 101–6. 2008, 168, 820–5. 48. Hoerger TJ, Segel JE, Gregg EW and Saaddine JB. 36. Tripathy D, Dhindsa S, Garg R, et al. Hypogo- Is glycemic control improving in the U.S. adults? nadotropic hypogonadism in erectile dysfunction Diabetes Care 2008, 31, 81–6. 34 Further Initiatives to Enhance Diabetes Care in Older People

Alan Sinclair Bedfordshire and Hertfordshire Postgraduate Medical School, University of Bedfordshire, Luton, UK

especially in the areas of assessment and treatment Key Messages [1]. This may be particularly important in older patients with diabetes who may have considerable (but • Improving diabetes care for older people is often undetected) impaired lower-extremity function. a major challenge for modern-day clinicians, The wide spectrum of vascular complications, acute and involves achieving high-quality care and metabolic decompensation, adverse effects of medi- standards across all clinical domains. cation, and the effects of the condition on nutrition • Inpatient diabetes care is a relatively understud- and lifestyle behaviour, may all create varying lev- ied area of research. However, the application of els of impairment and/or disability. These changes sound principles of assessment and prevention can do much to alleviate inpatient problems and may have adverse rebound effects on vulnerability to improve discharge outcomes. other comorbidities, independence and quality of life. • The development of a robust minimum dataset Advancing age itself – even in the absence of spe- for patients with diabetes will increase the cific diagnosed conditions – is associated with disabil- likelihood of consistency of outcome data, and ity, suggesting that disease prevention or amelioration provide a more meaningful interpretation of the would only be partially effective. results of clinical trials. Each disability has the potential to disadvantage individuals considerably (handicap), such as failure to enjoy outside entertainment and leisure activities, 34.1 Introduction and an inability to go shopping. Yet, handicap is not an inevitable occurrence, as many factors such as the One of the greatest challenges that the medical profes- reversibility of the intrinsic impairment, the presence sion faces in the twenty-first century is the epidemic of other medical comorbidities, mood, and even social relating to obesity and the marked rise in type 2 dia- support and financial status, can have dramatic effects betes. All members of the profession must honour their on the level of impact of the disability. own commitment to strive for excellence in medical Whilst the need for specific rehabilitation pro- care, and diabetes in ageing subjects poses one of the grammes for diabetes-related disability requires justi- more distinct of these challenges. fication, those clinicians managing older subjects will Thoroughness and vigilance are prime qualities that require a detailed knowledge of any assessments and are needed in managing older people with diabetes, available therapies via a multidisciplinary environment.

Diabetes in Old Age Third Edition Edited by Alan J. Sinclair © 2009 John Wiley & Sons Ltd. ISBN: 978-0-470-06562-4 466 CH 34 FURTHER INITIATIVES TO ENHANCE DIABETES CARE IN OLDER PEOPLE Also required is an ability to set not only the goals and often made for the first time at admission into hos- a realistic time frame for rehabilitation, but also what pital, since some 40% of older adults presenting with aides and appliances are available to achieve this. In hyperglycaemia have no previous history of diabetes. a similar manner to educational programmes, encour- In this situation it is imperative that the treatment of aging subjects to take an active part in rehabilitation diabetes is not delayed, as mortality may be significant can, in the present author’s experience, foster auton- within the first two days of admission. omy, improve self-esteem and coping skills and also Older adults pose several additional problems of reduce anxiety and depression. diabetes care relating to the time of admission, their in- Although this is complex area, several research patient stay, and the predischarge and discharge phases. groups have undertaken studies which take us forward Preventive strategies in the community to reduce hos- to understanding the nature of the disabling process, pitalization include the prompt treatment of urinary and and are continuing to pave the way for re-enablement chest and skin infections, the opportunistic screening in diabetes [2, 3]. Whilst the ‘multifactorial’ nature for diabetes in housebound elderly and care home res- of this process may prevent straightforward interven- idents [4], where early treatment will reduce metabolic tions from being effective, there is a need for a greater decompensation, and the early recognition of depres- understanding of the role of glycaemic control, and sive illness [5]. larger-scale randomized controlled trials need to be In the following section, further commentary on conducted to assist in this exploration. inpatient diabetes care will be provided to complement In addition, there is a need to stimulate the interest the detailed and comprehensive information on acute of clinical and laboratory research groups to provide hospital admissions in Chapter 24. These studies have the evidence that can justify particular therapeutic informed the basis of a contribution to a UK document terventions and promote specific patterns of care char- on emergency and inpatient diabetes care [6]. The basic acterized by three themes: (i) a major emphasis on the aims of inpatient care are listed in Table 34.1. quality of life and well-being of each patient; (ii) early and effective interventions; and (iii) a commitment to improve, or at least maintain, functional status. 34.2.1 Admission Acquiring unique knowledge and skills in geriatric Decompensation due to ketoacidosis is often rapid in diabetes enhances diabetes care for the vulnerable onset, whilst hyperglycaemic non-ketotic coma may directly, and often indirectly by influencing attitudes (avoiding ageism and a reductionist approach) to care. This book has been devoted to this goal. Table 34.1 Inpatient care of older adults with diabetes: The following two areas are distinct – but the aims of care. complementary – themes which have significant • Treat infection promptly and aim to stabilize glycaemia potential to improve diabetes care for older people: within 48 h of admission. • Improving inpatient diabetes care • To exclude nutritional impairment and ensure dietary • planning: consider enteral feeding early for those with The use of a diabetes minimum dataset (MDS) to critical illness or postoperative cases. enhance care. • To undertake a full functional assessment, including the assessment of mood, cognition, gait and socioeconomic 34.2 Improving inpatient diabetes factors. care • To diagnose all cases of diabetes in those presenting with acute hyperglycaemia: use both a fasting glucose and Older adults with diabetes have a two- to fourfold in- an oral glucose tolerance test where necessary, and be crease in the risk of hospitalization, while factors such prepared to repeat at time of discharge. as a high likelihood of significant preadmission med- • Ensure that a comprehensive predischarge assessment is ical comorbidities and disability often result in poor carried out to minimize readmission and increase the clinical outcomes and a prolonged length of stay. Ma- overall quality of diabetes care. jor vascular episodes such as a stroke or myocardial • Maintain dignity and equity of care in all hospital depart- infarction are common causes of admission in older ments at all times. patients with diabetes. The diagnosis of diabetes is 34.2 IMPROVING INPATIENT DIABETES CARE 467 follow an insidious course of several days, associ- loss in the previous three months] are present in as ated with signs of dehydration, reduced consciousness many as 25% of cases, with this situation being more and other signs of hyperosmolality. Thiazide diuretics, prevalent in those aged ≥80 years. All older adults calcium-channel blockers and steroids have each been require a nutritional assessment to be made within 48 h associated with precipitating hyperglycaemic crises in of admission, with signs of anorexia in particularly elderly patients. being sought. Currently, a suitable instrument for this, Hyperglycaemic emergencies require prompt treat- which is recommended by NICE and BAPEN, is the ment in order to reduce disabling morbidity and prema- Malnutrition Universal Screening Tool (MUST) [7]. ture death. The residents of care homes are particularly This approach does not require the measurement of vulnerable, with admission due to metabolic decom- both height and weight, and has been correlated with pensation carrying a high mortality rate, presumably clinical outcome. due to multiple comorbidities and a delay in admission. Treatment involves adequate rehydration, intravenous 34.2.4 Tube-feeding insulin and also antibiotics where infection is considered an underlying precipitating factor. In severe cases, Older adults often have a significant reduced caloric there should be no hesitation in transferring patients to intake due to acute illness, to the effects of multiple a critical care unit for intensive monitoring. drugs, and to cognitive impairment. Dysphagia – both temporary and neurologically induced (e.g. by a stroke) – requires enteral tube feeding via a naso- 34.2.2 Postoperative care gastric tube or percutaneous endoscopic gastrostomy In order to minimize thromboembolic complications, (PEG). Patients will benefit from a well-planned infection rates and prolonged hospitalization, postoper- enteral formula (high concentrations of fructose, ative diabetes care should be supervised where possible fibre and monounsaturated fatty acids), which has by the diabetes team with a focus on skin and wound been shown to optimize the patient’s nutritional care by employing pressure-relieving support mat- status and reduce postprandial glucose excursions. tresses, a tailored nutritional therapy that ensures ade- A daily dietary provision should be at least 25–35 quate calories and protein intake to offset the effects of kcal kg−1 body weight, and this should reduce the surgery and catabolism, and early mobilization to en- likelihood of hypoglycaemia in those receiving hance lower-limb functional recovery. The glycaemic insulin. Tube-feeding for many older adults is affected targets should be set at HbA1c <7.5%, fasting glucose by changes in mental performance or mood status, and <8.0 mmol l−1, and random levels of <9 mmol l−1,to may prove to be a barrier to successful compliance. reduce infection rates, enhance wound repair and avoid metabolic decompensation. The consideration for in- 34.2.5 Inpatient functional assessment sulin therapy should be an early decision. In view of the increased likelihood of disability in those with diabetes who are aged ≥80 years, patients re- 34.2.3 Nutritional and feeding issues quire not only an assessment of lower-limb strength Nutritional impairment – if untreated – leads to in- (to exclude proximal myopathy) but also a foot exam- creased mortality, a greater length of hospital stay, ination to assess their neurovascular status. Observa- additional drug therapy, and a higher infection rate tion of the patient’s gait is essential to identify any due to a depressed immune function. This in essence remedial problems and to plan muscle-strengthening reflects a ‘failure to thrive’ state, and is often compli- exercises: this process promotes the maintenance of cated by both water and electrolyte disturbances. Older mobility, reduces the falls rate, and may also lessen people with diabetes thus require an early assessment urinary incontinence – which may be present in up to of nutrient needs by a qualified hospital dietician, con- 30% of older patients with diabetes. sideration for influenza vaccination (depending on the In addition, in view of the high incidence of depres- time of year), and a programme of nursing and med- sion (a significant predictor of admission into hospital ical care which minimizes dehydration, pressure sore of older people with diabetes) and cognitive impair- development and pain. ment, all older people require an evaluation for the At the time of admission, signs of malnutrition presence of these conditions. This can be carried out [body mass index (BMI) <18.5 and >10% weight using a Geriatric Depression Score (GDS) [8], and a 468 CH 34 FURTHER INITIATIVES TO ENHANCE DIABETES CARE IN OLDER PEOPLE Mini-Mental State Examination score [9] or a 2-minute • To allow a more detailed analysis of the validity, Clock Test [10]. Each of these tests may identify is- reliability and sensitivity of existing measures, and to sues which require further specialist investigation and promote the development of new measures suitable treatment. for studies in older people. • By systematic review procedures and meta-analyses 34.2.6 Predischarge and discharge of studies using a recognized minimum data set, At the predischarge period, a close liaison with the there will be an increased likelihood of demonstrat- family and/or carers – including referral to social ing both clinical and cost-effectiveness of a range of services – is recommended. Clear nutritional instruc- interventions. tions and meal planning guidelines should be provided Several diabetes datasets have also been developed and by an experienced nutritional health professional. A published. The European Commission were recently predischarge visit by a specialist diabetes nurse is of involved in the design of the DIABCARD, which con- paramount importance. sists of diabetes-related data and administrative data A clear medication schedule is essential, document- [12] and forms part of an initiative towards having ing clearly what changes to treatment have taken place. European healthcards (EU/G7 healthcards). The DIA- This is necessary because hospitalization is known to BCARD has the advantage of being compatible with increase the postdischarge rate of hypoglycaemia, pre- other datasets being developed, such as the German Di- sumably due to inpatient discontinuance of usual ther- abetes Passport, V8. The Australian government have apy (e.g. a sulphonylurea), followed by the reinstitution produced a summary core dataset for diabetes which is of therapy which may create misunderstandings among primarily clinically orientated [13]. This has 37 items health professional staff, patients and families. for collection, ranging from demographics through to Those older adults with diabetes who require sub- procedure coding, and is designed to bring about a de- sequent admission (or return) to a care home should gree of national consistency in recording information receive a preadmission comprehensive geriatric assess- about people with diabetes. In 2003, a Scottish Dia- ment. Here, the patient’s further nutritional needs are betes Core Dataset was prepared for the collection of assessed, the feeding regimens planned, drug schedules data about people with diabetes in primary care [14]. rationalized, and specialist diabetes follow-up arranged This was a very comprehensive dataset (over 50 sep- as appropriate. Liaison with Primary Care is essential arate items, many of which have subcategories), but in these circumstances. has already required an extension due to new clinical contract arrangements for general practitioners in the 34.3 The use of a diabetes minimum United Kingdom [15]. dataset (MDS) to enhance care 34.3.1 Use of a MDS in a clinic (outpatient department): the first-stage As a means of enhancing the validity and consistency of data collected during clinical trials with older sub- appraisal process jects, a geriatric minimum dataset has been developed In the frailty model of diabetes [16], a framework was [11]. This approach ensures that the essential data are developed that provides further assistance with clinical collected to allow a more informed interpretation of decisions, and allows the scope for establishing mini- outcome data. It is reasonable to conclude that this ap- mum core assessment criteria. Previously, it has been proach may be valid in different disease areas, where recommended that the annual review process should a more uniform approach to research is required. now include an assessment of basic measures of activ- In the field of type 2 diabetes, a minimum data set ities of daily life (ADL) function, such as a Barthel test; might serve four primary purposes: tests of cognitive function, such as the Mini-Mental State Examination (MMSE) or Clock Test; a screen for • To provide a standardized method of assessment and depression, such as the Geriatric Depression Score; and outcome measures for conducting large-scale inter- an assessment of gait and balance, which can be simply vention studies with a randomized controlled design. estimated by the timed ‘Get Up and Go’ test (for de- • To enable valid comparisons of research findings in tails of all tests, see Ref. [16]). This integrated process different populations of patients. of comprehensive geriatric assessment can suitably be 34.3 THE USE OF A DIABETES MINIMUM DATASET (MDS) TO ENHANCE CARE 469 applied to diabetes, and allows investigators to create and metabolic control, with the other dimensions of the culture of defining criteria not only in an academic the MDS relating to the organization and delivery scenario but also in a clinical setting. of care and patient-centred outcomes/informal carer Several clinical intervention models have been pro- issues [22]. In some cases, these might also be used posed [17] which might also serve as a template for as prospective and longitudinal outcome data. determining the items of a diabetes MDS for interven- By a strict appraisal of the evidence-base for stud- tion research. These were developed for frail subjects ies in geriatric diabetes, the European Guidelines was with diabetes, and designed to detect early vascular also able to define up to 30 new research areas for complications, optimize functional status and improve future study. These areas naturally influence the com- well-being – the very ‘hallmarks’ of assessing diabetes ponents of a minimum data set, thus emphasizing the interventions. The models proposed were the metabolic need for flexibility and adaptation as a key property model, vascular model and the rehabilitation model. of a modern research MDS. With this background, The latter two would require a major input from mul- multiple potential areas for research by a randomized tidisciplinary (MD) staff. controlled design (RCT) in geriatric diabetes begin to Both, assessment and outcome criteria have pre- emerge. As an example, a priority list of research viously been published for diabetic residents of care trials in older subjects with type 2 diabetes might homes in the United Kingdom [18, 19]. In the British include: Diabetic Association Report, at assessment, 11 items were recommended which comprised a clinical data • Benefits of lifestyle intervention and/or therapeutic and medication list, BMI, blood pressure (BP), fun- approaches (e.g. ACE-inhibitor, insulin sensitizer, doscopy, feet examination, urine/blood tests and a di- etc.) in reducing the incidence of type 2 diabetes etary plan. The outcome data related to HbA1c, lipids in hypertension or other cardiovascular risk factors. and BP, nutritional status, frequency and severity of • Outcome of intensive treatment with oral agents hypoglycaemia, frequency and outcome of hospital ad- and/or insulin in reducing primary macrovascular missions, vascular complication rate, change in level and microvascular outcomes, and mortality. of dependency and mental function, and quality of life/mobility measures. These can be extrapolated eas- • Benefits (vascular/mortality outcome data; ily to intervention studies in older people with type cost-effectiveness) of statin and/or fibrate therapy in 2 diabetes, depending on the nature of the interven- proven cardiovascular disease. tion, for example, a drug intervention or care model • Benefits of prolonged comprehensive geriatric as- approach. sessment programmes (CGA) (>12 months) in de- The Minimum Data Set/Resident Assessment Instru- termining clinical outcomes and longevity. ment (MDS/RAI) was developed in response to a series of scandals relating to the quality of care in long-term • Does lowering blood pressure reduce the risk of facilities in the United States [20]. This is a comput- dementia in type 2 diabetes and hypertension? erized system designed to be used by nursing staff • Value of educational approaches in the prevention in the development of care plans. It has 17 domains of diabetic foot disease in terms of behaviour and in the assessment phase and 18 separate resident assessment protocols dealing with issues such as acute knowledge, frequency of ulceration and amputation, confusional states, visual function, mood, falls, nutri- quality of life and patient satisfaction. tional status, and so on. While not directly categorizing • Value of both clinical and educational approaches by disease, these items have direct relevance to resi- in reducing the severity of hospital admission and dents with diabetes. metabolic decompensation, infections, and pain out- A set of European-wide Clinical Guidelines [21] comes in diabetic residents of care homes. indicated assessment tools which could form part of the basis of defining a MDS based on functional status. The 34.3.2 A minimum data set for intervention latter item is one of the principal assessment/outcome trials in type 2 diabetes: defining the criteria for interventional studies in older people with purpose and content diabetes. The others relate to the assessment of subjects at the time of clinical trial entry, to outcome data The chosen will ideally need to satisfy four principal relating to the quality of life, health economic data criteria: 470 CH 34 FURTHER INITIATIVES TO ENHANCE DIABETES CARE IN OLDER PEOPLE

1. Each item should be directly applicable and have Table 34.2 The possible components of a diabetes clinically relevance to the patient with type 2 MDS. diabetes. Metabolic and laboratory 2. As far as possible, all functional, quality of life • Waist–hip ratio and waist circumference/BMI and patient-centred outcomes should have been val- • Glycosylated haemoglobin (HbA ) idated in older populations. 1c • Fasting glucose 3. Each item should retain validity over time, making • Postprandial glucose it suitable for prospective study designs. • Fasting insulin 4. Each item within the dataset should be precise, • Full lipid profile, including LDL- and HDL-cholesterol unambiguous and acceptable to all major research stakeholders. • Blood pressure • An agreed MDS-Diabetes would need to be considered Serum creatinine as the minimum dataset that research organizations • Microalbuminuria (whether public or private) should use to collect • Uric acid, C-reactive protein and interleukin-6, tumour consistent and standardized (unified) information necrosis factor-alpha about the value of interventions. In addition, a Clinical MDS-Diabetes (as mentioned above) can be Quality of life established from this, so that health care organizations • Euro-QOL can adopt these items for benchmarking and quality • ADDQOL Senior assurance purposes. It can also be used by health care staff as part of the clinical audit process. By Patient-centred outcomes providing this standardized framework, clinical • Hypoglycaemia rate services and networks providing diabetes care to • Hospital admission rate older people can be evaluated critically from a quality • Satisfaction with treatment: DSTQ perspective, and modifications of the dataset can be • used for Commissioning purposes both in primary Self-rated health and Health status: SF-36 and secondary care settings. Future evidenced-based • Pain control decision making will also be enhanced as these • Carer Strain Index datasets become uniformly employed. An analysis of studies involving older subjects with Other diabetes-specific data items diabetes from the electronic databases of Medline, • Cardiovascular event rate, stroke rate, fatal and non-fatal EMBASE and CINAHL, suggests that the spectrum • Myocardial infarction (MI) for macrovascular and hyper- on outcome measurement is often not extensive, tension intervention studies with common repeating sequences such as body • Any diabetes-related end-point, including erectile dys- weight, HbA1c, stroke rate, cardiovascular event function rate and adverse event rate. However, many studies lack depth of enquiry by restricting the number Cost-effectiveness/health economics of variables examined; this – combined with the • No diabetes-specific data sets identified recruitment of poorly documented subjects – can limit the effectiveness of any well-designed minimum data set. This must be borne in mind in collaborative to require items relating to mortality, inpatient care, efforts when developing a final MDS. quality of life and health economic analysis. For studies in diabetes, the extensive list of items It is also anticipated that data collection will need provided in Table 34.2 should be considered as impor- to include guidance on minimum criteria for enrolment tant elements of this disorder (disease)-specific MDS, in the clinical trial relating to referral criteria, demo- although some further selection will be necessary. graphics and the precise assessment of diabetes-related Outcome variables which measure the effectiveness complications and current treatments. This may re- of differing models of care are not generally avail- quire additional effort to define minimum datasets for able for studies in diabetes. They are, however, likely patient demographics, patient categories by diagnosis 34.4 CONCLUSIONS 471 including patients in hospital, outpatient clinics, and 8. Burke WJ, Roccaforte WH and Wengel SP. The short those in care homes, treatment categories and vascular form of the Geriatric Depression Scale: a comparison complications at entry to the trial. with the 30-item form. J Geriatr Psychiatry Neurol 1991; 4 (3): 173–8. 9. Folstein MF, Folstein SE and McHugh PR. 34.4 Conclusions Mini-mental state. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr In order for progress to be made, an international agree- Res 1975; 12 (3): 189–98. ment is required to collect this specified minimum 10. Shulman KI. Clock-drawing: is it the ideal cognitive dataset in future clinical trials. This requires coopera- screening test? Int J Geriatr Psychiatry 2000; 15 (6): tion between all key organizations involved in diabetes 548–61. care and those involved in the specialist care of older 11. Abellan Van Kan G, Sinclair A, Andrieu S, Olde people. A consensus on the final list of items and for- Rikkert M, Gambassi G and Vellas B. The Geriatric mat of the dataset is overdue and is now required. Minimum Data Set for Clinical Trials (GMDS). J Nutr Health Aging 2008; 12 (3): 197–200. Diabetes in the elderly represents an exciting phase 12. The DIABCARD Data Set. (2001) Version 4.0 EU/G7 of research because the special issues involved in the Administrative and Emergency Data. Contact: diab- effective management and realization of important [email protected]. gaps in care provision are being increasingly recog- 13. National Health Data Committee (now HDSC), Aus- nized [23]. Moreover, today an increasing number of tralian Institute of Health and Welfare. (2003) The Di- major clinical trials involve older subjects, thereby abetes (Clinical) Data Set Specifications 2003. ISBN 1 providing the data to support specific interventions, 74024 287 4. even among the aged. 14. Scottish Executive, Edinburgh. (2003) Scottish Diabetes Core Dataset December 2003. SCI Diabetes Collaboration and SCIMP. Available at: www. References diabetesinscotland.org and www.ceppc.org/scimp. 15. Quality Outcomes Framework, 2008/9. Department of 1. Sinclair AJ. Special considerations in older adults with Health. Available at: www.dh.gov.uk. diabetes: meeting the challenge. Diabetes Spectr 2006, 16. Sinclair AJ. Towards a minimum data set for interven- 19: 229–33. tion studies in type 2 diabetes in older people. J Nutr 2. Gregg EW, Beckles GL, Williamson DF, Leveille SG, Health Aging 2007; 11 (3): 289–93. Langlois JA, Engelgau MM and Narayan KM. Dia- 17. Sinclair AJ. Diabetes in old age – changing concepts betes and physical disability among older U.S. adults. in the secondary care arena. J Roy Coll Phys (Lond) Diabetes Care 2000; 23 (9): 1272–7. 2000; 34 (3): 240–4. 3. Volpato S, Blaum C, Resnick H, Ferrucci L, Fried 18. British Diabetic Association. (1999) Guidelines of LP and Guralnik JM. Comorbidities and impairments practice for residents with diabetes in care homes. explaining the association between diabetes and lower British Diabetic Association, London. extremity disability: The Women’s Health and Aging 19. Sinclair AJ, Turnbull CJ and Croxson SCM. Document Study. Diabetes Care 2002; 25 (4): 678–83. of Diabetes Care for Residential and Nursing Homes. 4. Sinclair AJ and Gadsby R. Diabetes in Care Homes. Postgrad Med J 1997; 73: 611–612. In: A.J. Sinclair and P. Finucane (eds), Diabetes in 20. Hawes C, Morris J, Phillips C, Fries B, Murphy K and Old Age, 2nd edition. John Wiley & Sons, Chichester, Mor V. Development of the nursing home Residents 2001; pp. 241–252. Assessment Instrument in the USA. Age Ageing 1997; 5. Rosenthal MJ, Fajardo M, Gilmore S, Morley JE and 27 (Suppl. 2): 19–25. Naliboff BD. Hospitalisation and mortality of diabetes 21. European Diabetes Working Party for Older People in older adults. A 3-year prospective study. Diabetes 2001–2004: Clinical guidelines for type 2 diabetes mel- Care 1998; 21 (2): 231–5. litus [article online]. Available from www.eugms.org. 6. National Diabetes Support Team (2008) Improving 22. Sinclair AJ, Turnbull CJ and Croxson SCM. Document emergency and inpatient care for people with diabetes. of care for older people with diabetes. Postgrad Med J Available from [email protected]. 1996; 72: 334–8. 7. Stratton RJ, King CL, Stroud MA, Jackson AA and Elia 23. Sinclair AJ. Aging and diabetes. (2004) In: Inter- M. ‘Malnutrition Universal Screening Test’ predicts national Textbook of Diabetes Mellitus, 3rd edition. mortality and length of hospital stay in acutely ill R.A. De Fronzo, E. Ferrannini, H. Keen and P. Zim- elderly. Br J Nutr 2006; 95 (2): 325–30. met (eds). John Wiley & Sons, Chichester, UK, pp. 1579–97. Index

Note: Page references in italics refer to Figures; those in bold refer to Tables abscess 124 anorexia–sarcopenia–sachexia triad bipolar I disorder 16 Absidia 341 210–11 bladder dysfunction 144 acarbose 249–50 anticonvulsants in diabetic neuropathy blood glucose meters 28–9 acetohexamide 246, 247 145 body composition changes 210 acid–base balance 199–200 anti-diabetic agent, choice of 257–9 body mass index (BMI) 14 adiponectin 5, 185 anti-epileptics 114 body weight 14 adult respiratory distress syndrome antipsychotic mediation 15–16 buformin 248 204 appetite 210 bunions 114, 139 admission plasma glucose (APG) 28 aripiprazole 16 advanced glycosylation end product artificial nutrition 223–5 (AGE) levels 44, 45, 109 aspirin 47, 87, 176 C-reactive protein (CRP) 5 adverse drug reactions 315 Aß classification 32, 33 calcium 221 AERx 256 atenolol 59, 84, 85, 276 calcium channel blockers (CCBs) 58, 59, age as risk factor of diabetes 13 atherogenic lipoprotein profile 169 97, 276 AIR 256 atherosclerosis 42–4, 43 callus formation 118, 123, 139 albumin: creatinine ratio (ACR) 94 atorvastin 283, 284 Cambridge Risk Score 31 alcohol 4, 220 atypical antipsychotics 32 candidiasis diabetic foot and 119 Auditory Verbal Learning Test (AVLT) oesophageal 341 hypoglycaemia and 301 374 oral 341 aldose reductase 109 auto-amputation 128 capdaicin 114 allodynia 138, 140 autonomic diarrhoea 143–4 captopril 84 alpha-blockers 97 autonomic neuropathy 142–4, 406 carbamazepine 145 alpha-glucosidase inhibitors 79, 96, 245, cardiovascular 142–3 carbohydrate metabolism 4 249–50 clinical consequences 142 carbohydrates 218–23 alpha-lipoic acid 145 gastrointestinal 143–4 carbutamide 246 alprostadil 157, 158 autonomic symptoms 289 cardiovascular autonomic neuropathy alteplase 80 auto-sympathectomy 118 142–3 Alzheimer’s disease (AD) 75, 173, 228, cardiovascular autonomic function tests 364, 374, 460 143 aminoglycosides 341–2 balance 407, 411 changes in peripheral blood flow 143 amitriptyline 144 Barthel Activities of Daily Living (ADL) postural hypotension 142–3 amlodipine 85, 276 scale 237, 242, 330 cardiovascular disease (CVD) 51 ampicillin 125, 144 Beck Depression Inventory (BDI) metabolic syndrome 54–5, 55 amputation 392 prevalence 53 foot 128–9 Benton Visual Retention Test (BVRT) as risk factor 15 Lisfranc 129 374 cardiovascular function tests, reference lower-limb 113–14 Berger’s disease 45 values 143 ray 125, 125 beta-blockers 32, 58, 84, 97, 154, 275 cardiovascular risks 53–61 amylin analogues 253–5 GIK therapy 62 absolute risk 53 anaemia 98, 407 in heart failure 63–4 absolute risk reduction 54 angiotensin receptor blockers (ARBs) 58, myocardial infarction and 62 behaviour modification 54 59, 97 beta-cell failure 6 dyslipidaemia 56–7 angiotensin-converting enzyme (ACE) beta-receptor agonists 32 hypercoagulability 60 inhibitors 6, 58–9, 84, 88, 97, 176, bezafibrate 175 hyperglycaemia 59–60 275 bicarbonate therapy 203–4 hypertension 57–9 ankle-brachial index (ABI) 42, 43,47 biguanides 245, 248–9, 253 intervention 54 ankle pressure index 117 bioelectrical impedance analysis (BIA) metabolic syndrome 54–5, 55 anorexia 213, 228 216 number needed to treat (NNT) 54

race and ethnicity 60–1 cognitive dysfunction 373–81, 407, DSM-IV-R and ICD-10 386 relative risk 53 459–60 epidemiology 387–8 risk reduction 47, 54 detection 377–9, 377, 380 general population 387 cardiovascular system 51–3 influence on self-care 378–9 older population 387–8 effect of ageing 51–2 methods 377–8 maintenance therapy 394 effect of diabetes 52–3 diabetes and 374–6, 374 mild to moderate depression 394 care homes, diabetes in 311–22 exclusion of depression 380 in older people 387 care home resources 321–2 further investigations 380 pharmacotherapy 393–4 complications and comorbidity glycaemic control and 376–7 psychological interventions 394 313–14 cognitive functioning 172–3, 188, 373 screening 390, 391 epidemiology 312 communication difficulties 315 stress-vulnerability model 389–90 local level resources 322 community care see primary care stresses associated with diabetes 390 management problems 314–15, 314 community matrons 332 treatment 393–4 organization of diabetes care in 315–21 community pharmacist 423–4 depression care management 394 assessing efficacy and efficiency of comprehensive geriatric assessment depression case management 394 321 188–9, 190 Dermagraft 124 developing standards 315–18 computed tomographic angiography 47 DESMOND 435–6, 440 eye care 320 computed tomography of diabetic foot detection of diabetes, reasons for 23–4 footcare 315, 320 124 diabetes care, inpatient 466–8, 467 improving care 318 congestive heart failure (CHF) 62 admission 466–7 multidisciplinary diabetes care 319 constipation 220 diabetes minimum dataset (MDS) nursing care 319–20 copper 222 468–71 nutrition 314, 318 Cornell Scale for Depression in Dementia in clinic (outpatient department) responsibility of physician 318–19 (CSDD) 392 468–9 setting standards 315 coronary artery bypass grafting (CABG) components 470 UK National Service Framework 358 intervention trials 469–71 Standards 315, 316 coronary heart disease 51–66 inpatient functional assessment 467–8 prevalence 313 incidence 52 nutritional and feeding issues 467 sustaining 321–2 epidemiology 53 postoperative care 467 in the UK 311–12 anti-diabetic medications and 64–6, 66 predischarge and discharge 468 carers see also cardiovascular disease; tube-feeding 467 benefits of 449, 449 cardiovascular risks diabetes education 427–41 effect of caring on 447–8 coronary risk equivalent, diabetes as 51 adaptations of educational principles needs 448–9, 448 corticosteroids 185 for the elderly 437–9 role 445–6, 446 cortisol 8 adult, principles of 428 support of 443–9 cranial mononeuropathies 141 care planning model 431 carpal tunnel syndrome 141 critical ischaemia 44, 46 carvedilol 59, 64 Cystatin C 94 choice of insulin 439–40 central obesity 32 cystitis, emphysematous 341 coping with illness 440 Centre for Epidemiological Studies cytokines 109 delivery 428–9 Depression scale (CES-D) 392 development 434–7 cerebrovascular disease as risk factor 15 empowerment in diabetes care Charcot feet (neuroarthropathy) 120, DASH diet 218 429–34 121, 139, 143 deadly quartet see metabolic syndrome empowerment in the elderly 431–2 cheiroarthropathy 118 definition of diabetes 21–3 barriers to 432–3 Chlamydia 341 dementia 16 feelings about 434 chlorpropamide 78, 96, 197, 238, 246, depression measures in 392 group 428–9 247, 301–2 vascular (VaD) 75, 374 hypoglycaemia 440 chlorthalidone 276 see also Alzheimer’s disease individual assessment 439 cholecalciferol 98 depression 16, 188, 368, 385–97, 407, insulin refusal 439 chromium 4, 222 460 Joslin Clinic principles 428 chronic kidney disease (CKD) 95 aetiology 389–90 one-to-one vs group 429, 430 classification 95 assessment 390–2 polypharmacy 440–1 criteria for referral 96 aim 390 diabetes educator 422–3 cilostazol 48 in dementia 392 diabetes prediction calculators 31 ciprofloxacin 125 general population measures 392 Diabetes Problem-Solving (Toobert and clawed toes 114 in primary care 390–1 Glasgow) 379 clindamycin 123, 126 process 390–1 Diabetes Specialist Nurses (DSNs) 319, Clock in Box test 379 associated problems 388–9 420, 422–3 clock-drawing test (CDT) 376 diabetes complications 389 diabetic amyotrophy 140–1 clonidine 154 health care expenditure 389 diabetic cardiomyopathy 53 clopidogrel 47, 87 metabolic control 388–9 diabetic foot 113–31, 139, 320 Clostridium 341 self-management 388 at-risk foot 120–1, 121 Clostridium difficile 227 case studies 395–7 autonomic neuropathy 116 clozapine 16 definition 385–7 bleeding 123 co-amoxiclav 123 diet and 213 in care homes 315 INDEX 475

classification (Wagner) of ulceration proteinuria 96 eligibility for lipid-lowering therapy 120 renal transplantation 98 284 deep ulcers: Wagner/UT Grades 2 and renin-angiotensin system (RAS) LDL-cholesterol level 282 3 124–6 blockade 97 randomized controlled trials 282–4 diabetic Charcot foot 129–31, 130Ð2 screening 94–5 risk assessment and treatment targets diet and 228–9 diabetic retinopathy 103 281 extensive gangrene: Wagner Grade 5 background 105 as risk factor for peripheral arterial 129 disc neovascularization (NVE) 106 disease 42 foot care education 121 exercise therapy 107–8 treatment of 281–4 foot pressure abnormalities 118–19 maculopathy 105 dysphagia 143, 228 infection 119, 123 management 104–6 limited joint mobility 118 nutritional treatment 107 localized gangrene: Wagner Grade 4 NVE elsewhere 107 Edinburgh Hypoglycaemia Scale 289 126–9, 126–7 pharmacological approaches 108–9 educational programmes 189 distal reconstructive operations pre-proliferative 106 Elderspeak 433–4 128–9 risk factors 103–4, 104 electrolytes 199–200 indications for revascularization screening 104, 236 emphysematous cystitis 341 127–8 surgical treatment 108 end-stage renal disease 98 proximal arterial reconstruction 128 treatments 106–7 enteral tube feeding 225–7 motor fibre loss 116 visual problems resulting from diabetes entrapment neuropathy 141 non-healing ulcer 123 mellitus 108 environmental factors 4 peripheral sensorimotor neuropathy diarrhoea 227 epinephrine 8 114–16 autonomic 143–4 erectile dysfunction 149–58 peripheral vascular disease 117–18 diastolic heart failure 63 afferent and efferent neuropathies 152, risk factors 119 diazoxide 304 152 history of previous foot problems 119 diet 4, 237 artery size hypothesis 152 slough 125 cardiovascular risk 54 CVD and 151–2 superficial ulcers: Wagner/UT Grade 1 in diabetic renal disease 97 definition 149 121–4, 122 ethical issues 228–9 depression and 152 diabetic ketoacidosis (DKA) 32, 198–9, diagnostic testing 153, 153 and food intake changes 210 198, 234, 235 drug use and 153 general recommendations 218 clinical presentation 200–1 first-line therapy 154–7 geriatric syndromes and nutrition 228 diagnosis 200–1 history of diagnosis 154 malnourished diabetic patients 218–19 laboratory findings 201 intracavernosal therapy 157 obese diabetic patients 219 pathogenesis 199–200 intraurethral therapy 157 dietary recommendations 218 diabetic nephropathy 93 as observable marker of diabetes dietician 423 diabetic neuropathy 137–46 mellitus 150 asymmetrical neuropathies 140 Digit Symbol Substitution (DSS) 374, organic vs psychogenic 153, 153 classification of diabetic 377 pathophysiology 151, 151 polyneuropathy 137, 138 digoxin 64 prevalence 149, 150 painful, management of 144–6 dimethylbiguanide see metformin risk factors 150 treatment algorithm 145, 146 dipeptidyl peptidase IV (DPIV) inhibitors second-line therapy 157–8 diabetic peripheral neuropathy (DPN) 7, 252–3 third-line therapy 158 137–40 dipyridamole 87 topical therapy 157–8 epidemiology 137 direct growth factor modulators 109 treatment 153–8 clinical assessment 139 disability 363–9 ergocalciferol 98 differential diagnosis 139, 139 functional 365–6 erythromycin 143, 144 acute painful neuropathies 139–40 in older adults: magnitude and ethical issues 228–9 of poor glycaemic control 140 character of problem 363–5 ethnicity 3–4, 14, 344–6 of rapid glycaemic control (insulin modifiable factors 366–7 as cardiovascular risk 60–1 neuritis) 140 risk reduction 367–8 epidemiology in the UK 344–5, 345 pathogenesis 142, 142 trends 368 futures 345–6 diabetic renal disease 93–100 district (community) nurses 319–20 improving access to services 345 blood pressure 97 diuretics 64, 97 stroke incidence and 76 conservative management 98–9 doctor, role of 424 European care model 453–8 diet and nutrition 97 domperidone 143 audit of quality of care 457–6, 457 dyslipidaemia 97 doxazosin 276 care organization 456–7, 457 end-stage renal disease and RRT 98 dressing, diabetic foot ulcers 124 Comprehensive Geriatric Assessment glycaemic control 96–7 drug therapy 189–91 (CGE) in frail subjects 455 haemodialysis 98 interdependence 189 control targets 454–5 management 95–7 mechanism of action 246 current clinical guidelines 454–6 management of cardiovascular risk 97 see also under names healthy ageing, frailty and pathological management of CKD 97–8 dual-energy X-ray analysis (DXA) 216 ageing 454 microalbuminuria 96 duloxetine 114, 145, 393 physician and staff training 456 peritoneal dialysis 98 dyslipidaemia 97, 151 screening 454, 456 primary prevention 95 as cardiovascular risk 56–7 specificity of institutions 455–6 476 INDEX

European care model (continued) neuropathy and 411 haemodialysis 98 therapeutic education 455 foot pressure map 119 hallux rigidities 118 exenatide 252–3, 304, 353, 461 foot ulceration see diabetic foot HbA1 29, 30, 31 exercise 54, 219 Extended ADL score 330 Fournier gangrene 341 HbA1c 29, 30, 76, 77 Exubera 256 fractures 221, 404, 408 hearing impairment 406 eye frailty 183–91 heart failure 63–4 disease 103–10 biology 184 diastolic HF 63 healthy, importance of 110 biology of late-life diabetes 185–7 with a normal ejection fraction eye care 320 impact on frailty 186 (HFNEF) 63 ezetimibe 283 impact of frailty on development of heart-burn 143 diabetes 186–7 hepatic glucose production 5, 5 in late life 184–5 homeostasis model assessment index Facial Recognition Test (FRT) 374 prevalence 185 (HOMA) 175 falls 187, 221, 403–14, 460 screening tools 184 homocysteine 42, 172 comorbid diseases and conditions 407, fructosamine 30–1 hospital admissions, acute 337–9, 338 411–12 functional ischemia 44 discharge planning 338–9 complications of 404–5 fusion proteins 109 factors affecting management 338, 339 definition 403–4 hydralazine 64 environmental interventions 412–13 hypercortisolaemia 199 epidemiology 404–5 gabapentin 114, 145 hyperglycaemia 23, 195 evaluation 408–10 gait disorders 407, 411 hypercoagulability as cardiovascular risk fall-related injury 408, 412 gastric electrical stimulation (GES) 143 60 fluid status 406, 411 gastrointestinal autonomic neuropathy hyperglycaemia 460 glycaemic status 410 143–4 as cardiovascular risk 59–60 history 409 gastroparesis 143, 226–7 enteral tube feeds 225–6 hospital or nursing facility settings 410, gemfibrozil 175 falls and 405 413 genetic factors 3–4 intermediate 23 laboratory evaluation and radiologcal Geriatric Depression Scale (GDS) 380, hyperglycaemic coma 23 imaging 410 390, 391–2, 391 hyperhomocysteinaemia 60 management of older adults 410–13, geriatric syndromes and diabetes 187–8 hyperosmolar coma 195 411 gestational diabetes mellitus (GDM) 14 hyperosmolar hyperglycaemic state medications 407–8, 412 gingivitis 229 (HHS) 198–9, 198, 338 multifactorial risk factors in aetiology ginkgo biloba 48 clinical presentation 200–1 405 glibenclamide (glyburide) 78, 96, 197, and diabetic ketoacidosis insulin overall health status 407 238, 246, 247–8, 252, 267, 301–2 regimes 337–8 pathophysiology of, and risk factors gliclazide 302 diagnosis 200–1 405–8, 405 glimepiride 246, 252, 302 laboratory findings 201 patient safety and quality of care 413 glinides 197 pathogenesis 199–200 physical examination 409–10 glipizide 96, 197, 246, 247 treatment 201–4 prevalence 404, 404 gliptins 304 hyperosmolar non-ketotic coma (HONK) role of diabetes 405 gliquidone 246 198, 199, 234, 235 screening 408–9 glitazones 353 precipitating factors 235 family history 14–15 glomerulosclerosis 93 hyperparathyroidism 97 family support 443–9 glucagon 8, 8, 303 hyperphosphataemia 97, 98 fasting plasma glucose (FPG) 23, 25–7, glucagon-like peptide-1 (GLP-1) hypertension 28 inhibitors 7, 252–3 antihypertensive therapy as part of fats 218–23 glucose counter-regulation 8–9 multifactorial intervention 278 femoral neuropathy (diabetic amyotrophy) glucose-dependent insulinotropic as cardiovascular risk 57–9 140–1 polypeptide (GIP) 7, 252 indications for treatment 274 fenofibrate 175 glucose disposal rates, insulin-mediated pharmacological therapy 278 fibrates 88, 175, 283 5–6, 6 randomized controlled trials 275–7 fibre 220 glucose effectiveness 7, 7 meta-analysis: old vs new drugs 275–6 Finger Tapping Test 374 glucose-induced insulin release 5, 5 prospective studies in old people Finnish Diabetes Risk Score 31 glucose-potassium-insulin (GKI) regimens 276–7 flucloxacillin 125 81–2 prospective studies in type 2 diabetic fludrocortisone 142 glucosteroids 32 patients 277 fluid balance 411 glutamic acid decarboxylase (GAD) 6 prospective studies: old vs new fluid therapy 202 glyburide see glibenclamide drugs 276 fluids 199–200 glycaemic control 96–7 risk assessment from observational fluoxetine 393 in diabetic neuropathy 144 studies 273–4 fluvastatin 176 glycaemic targets 239 as risk factor for diabetes 15 folate 222 glyco-oxidation 44 as risk factor for peripheral arterial Fontaine state 46 glycosuria 25 disease 42 foot care 47, 320 glycosylated haemoglobin (GHb) 29–30 target values 274–5, 274 foot disorders growth hormone 8 treatment of 273–8, 274 falls and 407 gustatory sweating 144 hypertriglyceridaemia 225 INDEX 477 hypocalcaemia 97 treatment in elderly people 340–1, 342 lispro insulin 255–6 hypochloraemic metabolic acidosis 204 inhaled insulin 256, 353 loop diuretics 7 hypoglycaema 195–7, 205, 270, insulin 202–3, 224, 255–7, 256 losartan 59 287–305, 338 assistive devices 190–1 Lys.Pro insulin 197 acquired hypoglycaemia syndromes in choice of 439–40 Type 1 diabetes 290–2 inhaled 256, 353 central autonomic failure 292 pulsatile release 4 macular edema 108 counter-regulatory deficiencies self-administration difficulties 191 magnesium 4, 222 290–1, 291 insulin analogues 255–7, 256 malaise, general 289 impaired awareness of insulin aspart 197 malnutrition 187–8, 216, 217 hypoglycaemia 291, 292 insulin aspart 256 pathophysiology 211 adjunctive therapies 304 insulin detemir 257 malondialdehyde (MDA) 44 adverse drug interactions 302, 302 insulin glargine 257 MAOIs 393 adverse effects 299–300 insulin glulisine 256 Meals on Wheels 216, 216 age and physiological response 292–4, insulin-induced hypoglycaemic coma 297 medial arterial calcification 114 293 insulin-mediated blood flow 5–6, 6 Medicated Urethral System for Erection cardiac sequelae 300, 300 insulin refusal 439 (MUSE) 157 in care homes 314 insulin resistance meglitinides 79, 245, 251–2 cognitive function 293–4, 293 and arterial hypertension 170–1 mental illness 15–16 counter-regulation 288–9, 288, 292, and dyslipidaemia 168–9 metabolic acidosis 98 294–5, 294 and obesity 169–70 metabolic alterations 5–7 education 440 insulin resistance syndrome see metabolic metabolic decompensation 195–205 epidemiology 296–9 syndrome metabolic syndrome 33, 54–5, 55, 79, earlier studies 296–7 insulin therapy 265–71 188, 238 frequency of hypoglycaemia 296 choice of 269 AACE clinical criteria 166 interventional studies 297–8, 297 favourable outcomes 271 as cardiovascular risk 54–5, 55 observational studies 298–9 glucose-powering medications and clinical 166–7 falls and 405 268–9 cognitive functioning 172–3 glycaemic thresholds 290, 290 glycaemic threshold as trigger for 266 definition 165–7 morbidity 299–300 indications 266–7 in high-risk CVD patients 176 mortality 299 introduction of complex regimens 271 insulin resistance nocturnal 196 special considerations 267–8 and arterial hypertension 170–1 neuropsychological manifestations 300, starting 270–1 and dyslipidaemia 168–9 300 titrating dosage and monitoring and obesity 169–70 oral antidiabetic agents 301–2 progress 270 homocysteine 172 physiological responses 288–90, 288 intermittent claudication 44, 47–8 in moderate-risk CVD patients 176 prevention 304–5, 305 intravitreal injections of trimacinolone muscle functioning 173 risk factors 300–2, 301 acetonide (IVTA) 109 NCEP ATP III clinical criteria 166 silent 196 iron, dietary 4, 222 in older persons 168, 171–3 sulphonylurea-induced hypoglycaemia ischaemic stroke 75–6 oxidative stress 171–2 303–4 isolated post-challenge hyperglycaemia pathogenesis 167–8 symptomatic response 196, 292–3, 293 22 potential components, in the elderly symptoms 289–90, 289, 295–6, 304 171–3 treatment 302–4, 303 therapeutic perspectives 173–6, 174 type 2 diabetes and responses 294–6 ketoacidosis 195 weight management 173–4 Hypoglycaemia-Associated Autonomic ketogenesis 199 WHO clinical criteria 166 Failure (HAAF) 292 ketosis-prone Type 2 diabetes mellitus metformin (dimethylbiguanide) 64–5, hypogonadism 460 32, 198 78–9, 96, 197, 224, 240, 248, 259, hypogonadotrophic hypogonadism 151 kidney, diabetic, changes in 93–4 304, 461 hypokalaemia 205 Kimmelstiel–Wilson lesions 94 methicillin-resistant Staphylococcus hypophosphataemia 204 Klebsiella 341 aureus (MRSA) 314 hypovolaemia 199 known diabetes 31–2 metoclopramide 143, 154 Kussmaul–Kien respiration 201 metronidazole 125 microalbuminuria 55, 94, 96, 237 imipramine 144 microcirculation 45 impaired fasting glucose (IFG) 14, 22 lacosamide in diabetic neuropathy 145 midodrine 143 impaired glucose tolerance (IGT) 6, 14, lactic acidosis 249 miglitol 249–50 22 lacunar stroke 75 minerals 220–3 incretin 4, 7, 245–6 latent autoimmune diabetes in adults Mini-Mental State Examination (MMSE) indapamide 85, 277 (LADA) 6, 32 237, 242, 330, 374, 375, 376, 377, infections 340–3, 340 leg ulceration 315 378, 379, 438 in care homes 314 Legionella pneumophilia 341 Mini Nutritional Assessment (MNA) management 341 lifestyle factors 4 211, 214–15 microbiologist, role of 342–3 lignocaine, intravenous 145 minimum dataset–resident assessment precautions 341–2 Lisfranc amputation 129 instrument (MRS-RAI) 321, rare 341 lisinopril 276 468–71, 470 478 INDEX minoxidil 158 obesity 4, 314 peripheral vascular disease (PVD) 113, mitiglinide 251 central 32 114 mobility 187, 407, 411 insulin resistance and 169–70 Peyronie’s disease 158 Modification of Diet in Renal Disease overnutrition and 213–17 phenformin (phenethyl biguanide) 248 (MDRD) formula 94, 95 sarcopenic 186 phentolamine 157 modified oral glucose tolerance test octreotide 304 phosphate therapy 204 (MOGTT) 27 oesophageal candidiasis 341 phosphodiesterase type 5 (PDE5) Modified Wisconsin Card Sorting Test olanzapine 16 inhibitors 154–7, 157 379 One Touch II meter 27 physical training 189 molecular biology studies 8 ophthalmoplegia 141 physical inactivity as risk factor 15 monoclonal antibodies 109 opiates 145 pioglitazone 65, 66, 79, 96, 174, 250, monofilaments 115 oral agents 254 251, 259, 353, morphine 145 oral candidiasis 341 461 oral glucose tolerance tests (OGTTs) 15, plexopathy (diabetic amyotrophy) 140 22, 23, 25–7, 33–4, 236 plurimetabolic syndrome see metabolic mortality, depression and 389 modified 27 syndrome motilin 143 oral health 229 podiatrist 423 Mucor 341 oral vasodilator prostaglandins 48 Podotrack 118 multidisciplinary team (MDT), role of osmotic symptoms 24 polydipsia 236 419–25 osteoarthritis 41 polymyositis 140 care home residents 425 osteomyelitis 123, 124, 126 polypharmacy 408, 412, 440–1 changing role 420 osteoporosis 221 polyuria 24, 236 ethnicity 425 otitis externa, malignant 341 postprandial hyperglycaemia (PPHG) key roles 422 outflow procedure 128 245 patient empowerment and role 420–1 overnutrition and obesity 213–17 postprandial plasma glucose 27–8 relationship with other services 420, prevalence 213–14 post-stroke hyperglycaemia (PSH) 421 weight loss advice 214 79–80 role in age-related frailty 424–5 oxidative stress 171–2 postural hypotension 406, 411 terminal illness 425 oxycodone 145 potassium therapy 203 muscle functioning 173 pramlintide 240, 246, 255, 461 Mycoplasma 341 pravastatin 331 myocardial infarction 51, 60–1 Paced Auditory Serial Addition test prayer sign 118, 118 (PASAT) 374 pregabalin 114, 145 painful-painless foot 138 pressure palsies 141 nateglinide 224, 251, 252 pancreatic polypeptide (PP) 9 pressure sores 228–9 National Adult Reading Test (NART) pancreatitis 32 pressure stress 315 375 papaverine 157, 158 pressure ulcers 460 necrotizing cellulitis 341 paraesthesiae 138 PressureStat 119 necrotizing fasciitis 341 parenteral nutrition 227–8 priapism 158 neostigmine 144 Parkinson’s syndrome 187 primary and community care 325–34 neuroglycopenic symptoms 289 parodontitis 187 debate on 331 neuropathic cachexia 140 paroxetine 394 definition 325 neuropathy 406–7 pathophysiology of diabetes 3–9 delivering appropriate diabetes care autonomic 406 penile prosthesis 158 331–3 peripheral 406–7 penile revascularization 158 integration of care into the neuropathy disability score (NDS) penile trauma 158 community 332–3 115–16 pentoxyifylline 48 optimizing care and reducing nitrates 64 percutaneous endoscopic gastrostomy hospital admissions 332 nitrendipine 277 (PEG) 224, 225 funding and organization of healthcare nitroglycerine 158 peridostigmine 144 325–6 non-insulin-mediated glucose uptake perindolol 85 move from hospital to community (NIMGU) 7 perindopril 59, 277 326–7 nortriptyline 393 peripheral arterial disease (PAD) 41–8 polypharmacy and medication nursing care homes, definition 312 clinical presentation 45–6 concordance 333–4 nutrition 189, 209–29 asymptomatic 45 primary care diabetes team 327 assessment 214–18 claudication 45–6 quality and outcomes framework in care homes 314, 318 critical limb ischemia 46 (QOF) scores 329 inpatient 467 diagnostic methods 46–7 quality diabetes clinical indicators see also diet anamnesis and physical assessment 327–8 nutritional oral supplements 223 46 routine care of people by primary care Nutritional Risk Screening (NRS) 215 vascular diagnostic techniques 46–7 diabetes team 327–8 nutritional screening, validated methods epidemiology 42 strategies to prevent CVD 328–31 215 mortality rate 41 applying evidence to independent nutritional status, normal ageing and pathophysiology 42–5 and frail older people 330 210–11 treatment 47–8 blood pressure-lowering 330 nutritional support 210–14, 228 peripheral neuropathy 406–7 cholesterol control 328–30 INDEX 479

exception reporting 330–1 sex steroids 5 special problems in the elderly 352 glycaemic control 328 Sexual Encounter Profile 157 type 2 diabetes preparations 353 primary prevention 328–30 sickle-cell disease 29 symptoms of diabetes 24 primary care practice nurses 319 sildenafil 155–6 syndrome X see metabolic syndrome proinsulin 4 simvastatin 176, 282 proprioceptive impairment 406 sitagliptin 240, 253, 353, 461 protein 219 sleep disorders 16 tadalafil 156 protein kinase C (PKC) inhibitors 109 smoking telmisartan 276 protein-energy malnutrition 211 cardiovascular risk 54 tertiary care, definition 325 proteinuria 94, 96 diabetic foot and 119 testosterone 5 proximal motor neuropathy (diabetic as risk factor for diabetes 16 tetracyclin 144 amyotrophy) 140 as risk factor for peripheral arterial thalassemia 29 pseudoclaudication 45 disease 42 thiazide diuretics 4, 32, 59, 97, 154 Pseudomonas aeruginosa 341 sodium intake 220 thiazolidinediones (TZDs) 65, 79, 96, ptosis 141 sodium valproate 145 174, 197, 239–40, 245, 250–1, 253, pyelonephritis 341 specialist nurses, role of 419–25 259, 460 spironolactone 64, 154 thienopyridines 47 Staphylococcus aureus 341 third cranial nerve palsy 141 quetiapine 16 Staphylococcus sp., foot ulcers and thirst 24 119 thoracoabdominal neuropathy 141 statins 86, 88, 175, 282–3 ticlopidine 47 race see ethnicity in dyslipidaemia 283 tolazamide 246, 247–8 ramipril 58, 85, 276 in lowering cholesterol 56 tolbutamide 197, 246, 304 random plasma glucose (RPG) 27–8 steroids 109 tooth loss 187, 188, 229 ray amputation 125, 125 Streptococcus 119, 341 torcetrapib 284 Reaven’s syndrome see metabolic Streptococcus pneumoniae 341 total contact cast (TCC) 122 syndrome stroke 75–88 total reactive antioxidant potentials receiver operator characteristics (ROC) control of DM and stroke risk 77–9 (TRAP) 44 curve 27, 27 definitions 75–6 Trail Making B (TMB) 374, 375, 379 removable cast walkers (RCW) 122 diabetes, hyperglycaemia and 76–9 tramadol 145 renal bone disease 97–8 diabetes, post-stroke hyperglycaemia transient ischaemic stroke (TIA) 75 renal replacement therapy (RRT) 98 and prognosis after acute stroke transmetatarsal amputation 129 renal transplantation 98 79–80, 80 Treatment and Teaching Programme renin–angiotensin–system (RAS) epidemiology 76 (TTP) 438 blockade 97, 276 glucose-potassium-insulin (GKI) tricyclic antidepressants 114, 142, 144, repaglinide 224, 251, 252 regimes 81–2 393 residential care, definition 312 prevention in Type 2 diabetes 84–7 troglitazone 250, 251 retinopathy 23 antiplatelet therapy 86–7 truncal radiculopathy 141 rheumatoid arthritis 119 dyslipidaemia 86 tube-feeding 467 rhinocerebral mucormycsosis 341 hypertension 84–6 tuberculosis 341 Rhizopus 341 treatment 80–4, 82Ð3 tumour necrosis factor-alpha (TNF-alpha) rimonabant 175 Stroop test 379 5 risoiglitazone 65 Subjective Memory Questionnaire Type 1 diabetes 32 risperidone 16 379 Type 2 diabetes 32, 233–42 rosiglitazone 79, 174, 250, 251, 353, 461 sulphonylureas 64, 78–9, 197, 245, acute presentation 235 rosuvastatin 176 246–8, 247, 253 aims 235–7, 235 roziglitazone 96 drugs interactions causing defining characteristics 233 hypoglycaemia 302 high-quality diabetes care 234–5, Summary of Diabetes Self-Care Activities 234 salbutamol 32 (SDSCA) 378, 379 hyperglycaemia, symptoms and signs sarcopenia 45, 460 surgery, management of 349–58 234 sarcopenic obesity 186 emergency surgery 357–8 indications for referral 237 schizoaffective disorder 16 implications for 350–1 individual diabetes care plan 241–2, schizophrenia 16 insulins 352–3 241 Scotch-cast boot 122, 123, 131 new diabetes therapies and 352–3 metabolic targeting 241–2, 242 secondary care, definition 325 open-heart surgery 358 prioritizing diabetes care 242, 242 secondary diabetes 32 potential risks 351–2 initial treatment 237–41 selective serotonin reuptake inhibitors practical management 353–8 insulin therapy 240–1 (SSRIs) 393 aims 353 non-acute presentation 235–7 sensory ataxia 138 in insulin-requiring diabetes 355–7 diagnosis 236 sensory impairment 405–6, 411 GKI infusion 355–6, 356 identifying responsibilities 237 serotonin noradrenaline reuptake separate-line 355, 356 screening for complications 236–7 inhibitors (SNRIs) non-insulin-requiring diabetes 354–5, oral agents 238–40 in diabetic neuropathy 144–5 355 physical activity and exercise 238 sertraline 393 preoperative assessment 353–4, 354 role of diabetes specialist nurse 240 serum creatinine 94 problems induced by 349–50 treatment for diet failures 237 480 INDEX types of diabetes 32 urinalysis 237 vitamin D deficiency 220–1 tyrosine kinase 8 urinary incontinence 187, 315, vitamin E 4, 44, 48 407 vitamins 220–3 vitreous haemorrhage 104, 108 UK Quality and Outcomes Framework Vulnerable Elders Survey (VES-13) 413 (QOF) 327, 328, 330 vardenafil 156–7 ulceration, foot see diabetic foot vascular dementia (VaD) 75, 374 ulnar nerve entrapment 141 vascular endothelial growth factor weight 210, 212, 238 undernutrition 187, 211–13 (VEGF) 109 cardiovascular risk 54 United State care model 459–61 vascular thrombosis 205 management 173–4 complications in diabetes 459 venlafaxine 393 WHO criteria 24 cognitive problems 459–60 vestibular impairment 406 WHO-5 WEll-Being Index 390, 391 depression 460 vibration perception threshold (VPT) falls 460 115 hyperglycaemia 460 vildagliptin 240 X-PERT programme 435, 436–7, 437, hypogonadism 460 vision, falls and 406 440 pressure ulcers 460 visual acuity (VA) screening 236 sarcopenia 460 vitamin A 222 drugs 461 vitamin B 222 zinc 4, 222 management of diabetes 460–1 vitamin B12 (cobalamin) 221–2 ziprasidone 16 prevalence of diabetes 459 vitamin C 4, 222–3 special issues 459–60, 460 vitamin C 4 Index compiled by Alison Musker